# Colleviate Insights: full text > Field notes from Colleviate, a full-service marketing agency in Mumbai, India: growth audits, brand and campaign craft, search and AI-answer visibility, share of search, and marketing measurement. Each article opens with a short answer. Canonical URLs are given with each article; please cite them. --- # Our own Search Console, first ninety days: what a new site really sees URL: https://colleviate.com/blog/our-own-search-console-first-ninety-days.html Topic: Method | Published: 2026-11-03 | Updated: 2026-11-03 Publisher: Colleviate (https://colleviate.com/) **Short answer.** In its first three months colleviate.com earned 1,116 Google impressions and 3 clicks, at an average position of about 54 on desktop. Nearly nine in ten impressions came from outside India. About 120 pages are indexed, with 3 waiting. Visibility is growing fast from a very low base, and almost all of it sits on page 5 or later, which is why clicks have not followed yet. Most advice about new sites is written from the outside. This is the inside view: our own numbers, including the unflattering ones, and the decisions they led to. This is a case study with a single subject: this website. We are not describing a client, and nothing below is modelled. The figures come from our Search Console exports and the Page indexing report on 2 November 2026, covering web search for colleviate.com over the previous three months. ## The numbers | Measure | Result | | Impressions, three months | 1,116 | | Clicks | 3 | | Average position, desktop | 53.6 | | Last 14 days against the 14 before | 754 impressions against 208 | | Impressions from India | 104, with all 3 clicks | | Impressions from the United States | 485 | | Pages indexed | about 120 | | Pages waiting to be indexed | 3 (2 crawled, 1 discovered) | The chart above shows daily impressions. They were close to zero until 22 September, hovered around 15 a day through early October, and reached about 90 on 29 October. The last day in the chart is lower because Search Console reports the most recent days incompletely. ## What the data says - A new site is invisible, then briefly visible, then visible and unclicked. The first weeks had almost no impressions. Once pages were indexed, impressions grew, but at an average position of 54 a page is shown on roughly the fifth results page, where almost nobody clicks. - Impressions arrive before clicks, and from the wrong places. Eighty-seven percent of impressions came from outside India, mostly the United States, while the audience we want is Indian brands. The queries are English marketing terms with no national intent, so the site is being found for general topics, not for what is local to us. - Indexing is not the problem. About 120 pages are indexed and 3 are waiting. The report also lists 8 pages as alternates of a page with the right canonical, 3 redirects and 1 blocked by an unauthorised request. Those are housekeeping, not a crisis. Our technical SEO checklist (https://colleviate.com/blog/technical-seo-checklist.html) covers each status and what to do about it. - The gap is rank, not reach. Our biggest topic by impressions, share of search, sits at an average position near 75. Five pages sit at positions 6 to 15 with no clicks. For those, the job is a better title and snippet. For the rest, it is depth and links, not more pages. ## What we are changing because of it - Fix the snippets on pages already near page one before writing anything new. A page at position 7 with 70 impressions and no clicks is a title and description problem. - Deepen the pages that rank 40 or lower for the topics we want, and link to them from every related page, instead of writing a sibling that competes with them. The method is in internal linking for topic clusters (https://colleviate.com/blog/internal-linking-for-topic-clusters.html). When we audited our own links, 23 of 118 articles had no in-body links pointing to them at all, and we fixed that before adding new content. - Publish pieces built on our own data , such as the Share of Search Index (https://colleviate.com/blog/share-of-search-index-india.html), because a page with evidence nobody else has is the only kind a young site can win with. - Check Search Console weekly, including the indexing report, so a rise in pages that are crawled but not indexed stops us publishing before it becomes a pattern. How to read the report is in Search Console diagnostics (https://colleviate.com/blog/search-console-diagnostics.html). ## What we cannot claim Impressions started rising on about 18 October, before our 28 October release of the larger share of search cluster, so we cannot say what caused the rise. Three clicks is too few to say anything about click-through rate. Rankings move daily, and an export is a photograph of one afternoon. We will publish the same table again after another month, whichever way it goes. ## What to take from it If your own site is new and the numbers look like ours, you are not failing. You are early. The useful questions are the ones above: is Google indexing the pages, where do they rank, and which handful are close enough to page one to be worth rewriting. If you want that read on your own site, we do it as part of a search and answer-engine review (https://colleviate.com/blog/seo-vs-aeo-2026.html). ## Sources - Colleviate Search Console exports, 2 November 2026 (Performance and Page indexing), Colleviate, 2 Nov 2026: https://colleviate.com/ - Performance report (Search Console Help), Google, undated: https://support.google.com/webmasters/answer/7576553 - Page indexing report (Search Console Help), Google, undated: https://support.google.com/webmasters/answer/7440203 ## Questions **How long does a new website take to get search traffic?** In our case about six weeks passed before daily impressions rose above a handful, and clicks had not followed after three months. Your timeline will differ, and Google gives no guaranteed schedule. **Why does Search Console show impressions but no clicks?** Because impressions count any appearance in results, including pages far down the list. At an average position near 54, a page sits around the fifth results page, where few people click. **What does "crawled, currently not indexed" mean?** Google visited the page and chose not to add it to its index. Google lists it as a status to watch. If the number grows while you publish, slow down and improve the pages before adding more. **Is this a client case study?** No. It uses our own site and our own Search Console data, with no client information. --- # FAQ schema after the rich result ended: what to keep and what to cut URL: https://colleviate.com/blog/faq-schema-after-the-rich-result-ended.html Topic: AI visibility | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Google stopped showing FAQ rich results in Search on 7 May 2026, and had already limited them to government and health sites in 2023. FAQPage markup is still valid, but it no longer earns a visible search feature. Keep FAQs that answer real questions in plain text on the page, and stop reporting FAQ markup as a result. For years an agency could point to FAQ markup as a visible win in search. That win is gone. What is left is the part that always mattered: whether the questions on the page are worth answering. ## What did Google actually remove? It happened in two steps. In August 2023 Google announced that FAQ rich results, the expandable questions that appeared under a search listing, would be shown only for well-known, authoritative government and health websites, and that How-to rich results would be limited to desktop, which Google ended altogether in September 2023. Then, in a deprecation notice added to its FAQ documentation in May 2026, Google said the feature would no longer appear in Google Search from 7 May 2026. So for an ordinary business site FAQ rich results have rarely appeared since August 2023, and for everyone there are none now. A valid FAQPage block still passes a validator, because it is still valid schema.org markup, but it no longer produces anything you can screenshot. ## Does FAQ markup still do anything? We have found no published evidence that it improves rankings, and none that it improves how often an AI answer names or cites a page. Google has not said it does either. What the markup does is restate, in a machine-readable form, text that is already on the page. That is cheap and harmless, which is the only honest case for keeping it. We keep it on our own pages for exactly that reason, and we say so here rather than pretend otherwise. Every question in our markup is also visible on the page with the same answer, which is the one rule that has not changed: markup must describe what a visitor can see. ## What matters more than the markup The questions. A good FAQ block is a list of the questions a buyer really asks before they contact you, each answered in the first sentence and expanded in the second. That helps a reader, and a page that states an answer plainly is easier for any system to quote. A weak one is five sales lines phrased as questions, and no markup rescues it. The test is simple: would the page be better with this block if the markup did not exist? ## An audit you can run this week - Match markup to the page. Every question and answer in the markup must be visible on the page, word for word or very close to it. - Cut the decoration. Remove questions nobody asks, such as “Why choose us?”, and answers that are advertising. - Stop counting it. If a report from you or your agency lists FAQ rich results as a win after May 2026, ask what it is measuring. - Put the effort where it still pays. Organization, Article, Breadcrumb and Product markup still describe things Google documents. See schema markup that earns results (https://colleviate.com/blog/schema-markup-that-earns-results.html) and how Organization markup works (https://colleviate.com/blog/organization-schema-and-sameas-for-agencies.html). - Write for the answer, not the feature. How AI answer engines choose their sources (https://colleviate.com/blog/how-ai-answer-engines-choose-sources.html) explains why a clear, sourced sentence is worth more than a block of code. Markup that depends on a search feature is a bet on someone else's product roadmap. Content that answers a real question has lasted through every change in that roadmap. ## Sources - FAQ structured data, documentation and changelog (deprecation notice dated 8 May 2026), Google Search Central, 2026: https://developers.google.com/search/docs/appearance/structured-data/faqpage - Changes to HowTo and FAQ rich results, Google Search Central blog, August 2023: https://developers.google.com/search/blog/2023/08/howto-faq-changes ## Questions **Did Google remove FAQ rich results?** Yes. Google limited FAQ rich results to authoritative government and health sites in August 2023, and its documentation says the feature no longer appears in Google Search from 7 May 2026. **Should I delete FAQPage schema from my site?** You do not have to. It is still valid markup and costs little. Keep it only where every question and answer is visible on the page, and do not expect a search feature or a ranking benefit from it. **Does FAQ schema help AI answers?** We have found no published evidence that it does. A clear, sourced answer in the visible text is the better investment. --- # How to track whether AI answers name your brand URL: https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html Topic: AI visibility | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Track AI visibility as a rate, not a rank. Write 20 to 30 questions your buyers ask, run each one many times on each engine, and record how often your brand is named and how often your page is cited. A reported SparkToro experiment found the same prompt returns a different brand list almost every time, so a single run proves nothing. Ask an AI assistant who the best agencies are and you will get a list. Ask again and you will get a different one. Measuring that honestly takes more than one question. ## Why a single answer tells you nothing SparkToro, working with Gumshoe.ai, tested this. As reported by Search Engine Journal, the team ran 12 prompts, such as which chef's knife or headphones to buy, between 60 and 100 times each on ChatGPT, Claude and Google's AI. The same list of brands almost never came back twice, and the order was effectively random. The brands that led a category did keep turning up across runs, which is the useful part: the signal is how often you appear, not where. We are quoting a secondary report of that study, not its raw data, and your categories will differ. The method that follows does not depend on the exact numbers. It depends on the one point the reports agree on: one answer is one draw from a distribution. ## Why it is worth measuring at all When an answer is read and not clicked, being named is the only visibility left. Pew Research Center found that people clicked a traditional search result in 8 percent of visits that showed an AI summary, against 15 percent without one, and clicked links inside the summary in 1 percent. That is the argument for measuring citation, and for measuring it as carefully as share of search (https://colleviate.com/blog/how-to-measure-share-of-search.html). ## Step 1: write the question set Twenty to thirty questions, in the words a buyer would use, in four groups: - Category: “best marketing agency for a B2B brand in Mumbai”. - Comparison: “agency retainer or project fee”. - Problem: “why has our paid media stopped working”. - Brand: “what does your brand do”, to catch errors. Take the wording from real conversations (sales calls, enquiry emails, the questions customers actually asked), as customer research on a small budget (https://colleviate.com/blog/customer-research-on-a-small-budget.html) shows, not from your keyword list. ## Step 2: run each question many times Run every question at least 10 times per engine per month, in fresh sessions, logged out where the tool allows, with the same location and settings each month. For each run record: whether your brand is named, whether a page of yours is cited, whether what was said about you is right, and which competitors appeared. Note the date and the model name shown. ## Step 3: read it as a rate, with its margin Your mention rate is the runs that named you divided by the runs. That number has a margin of error, and it is wide at small samples. For a true rate of 30 percent, 10 runs give a reading that can swing by roughly 28 points either way, 40 runs by about 14, and 100 runs by about 9. So a move from 30 to 40 percent on 10 runs is noise. Only report a change when it is larger than the margin for your sample, and run more for the questions that matter most. | Runs per question | Margin on a 30% reading, 95% confidence | | 10 | about 28 points | | 40 | about 14 points | | 100 | about 9 points | ## Step 4: report it next to what you already track Put the mention rate beside your share of search and your share of voice, against the same named competitors, as in share of search vs share of voice (https://colleviate.com/blog/share-of-search-vs-share-of-voice.html). If you appear more often in AI answers than your search share suggests, the content is working. If less, start with why AI answers do not mention your brand (https://colleviate.com/blog/why-ai-answers-dont-mention-your-brand.html). Review it monthly, as part of a reporting cadence (https://colleviate.com/blog/marketing-reporting-cadence.html) the board already reads, and in long-consideration categories such as the ones in automotive marketing in India (https://colleviate.com/blog/automotive-marketing-india.html) look at it over quarters, not weeks. ## If you buy a tool Ask the vendor how many times each prompt is run per engine and how the margin is shown. A dashboard that reports a single position for your brand in an AI answer is reporting a number the research above treats as noise. ## Sources - AI recommendations change with nearly every query: SparkToro (secondary report), Search Engine Journal, 2026: https://www.searchenginejournal.com/ai-recommendations-change-with-nearly-every-query-sparktoro/566242/ - Google users are less likely to click on links when an AI summary appears in the results, Pew Research Center, 22 July 2025: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/ - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **How do you measure AI visibility?** As a rate: run a fixed set of buyer questions many times on each engine and record how often your brand is named and how often your page is cited, then compare with competitors and with your share of search. **Can you track your ranking in ChatGPT?** Not meaningfully. A reported SparkToro study found the order of brands in AI answers is effectively random, though the brands that lead a category appear often. Track how often you appear, not where. **How many times should I run each question?** At least 10 per engine per month, and 40 or more for questions that matter. At 10 runs, a 30 percent reading can swing by about 28 points either way, so small changes are noise. --- # Why AI answers do not mention your brand: six checks, in order URL: https://colleviate.com/blog/why-ai-answers-dont-mention-your-brand.html Topic: AI visibility | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Usually for one of six reasons: no page states plainly what you do, crawlers cannot reach your pages, nobody independent says it either, your name collides with something else, you do not match the question asked, or you are being asked in a language you have not covered. Each has a cheap test; run them in this order. When a buyer asks an assistant who to hire and your name is not in the answer, the cause is usually ordinary and fixable. Check the cheap things first. First, measure honestly. One answer that leaves you out is anecdote; a low rate over many runs is a finding, which is why the method in how to track whether AI answers name your brand (https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html) comes first. Then work down this list. We have not seen published evidence that ranks these causes, so treat each as a hypothesis to test with your own question set, not as a law. ## 1. Does a page say it in one plain sentence? An answer can only repeat what it can find. Write the sentence a buyer would want quoted, such as “Colleviate is a full-service marketing agency in Mumbai”, and put it on your home page and your about page, near the top, in text, not in an image. Test: paste that sentence into an assistant and ask it to summarise what you do. If it cannot, the sentence is not where machines can read it. ## 2. Can crawlers reach the page? Check your robots.txt, and the bot rules in your CDN, firewall and security plugin. Some of these block unfamiliar crawlers by default, and a page behind a login, a pop-up wall or a script that renders the text late may be invisible to a system that reads the HTML. If you cannot tell, ask your developer to fetch the page as a plain request and read what comes back. ## 3. Does anyone independent say it? A claim on your own site is a claim. The research paper that named generative engine optimisation, by Aggarwal and colleagues at KDD 2024, reports that content changes can boost a source's visibility in generated answers by up to 40 percent on their benchmark, with adding statistics and quotations among the strongest changes and keyword stuffing among the weakest. The practical reading is to publish facts someone can check, and to earn mentions from outlets that are not you. Where a fact is yours to own, publish the data: our Share of Search Index for India (https://colleviate.com/blog/share-of-search-index-india.html) exists partly for this reason. ## 4. Does your name collide with something else? This one catches many brands. When we built the index, a stockbroker called Dhan had to be left out because the word is also Hindi for wealth, and searches for Mahindra also count Kotak Mahindra Bank. If your name is a common word, or sits inside a bigger name, an assistant has to guess which you mean. Fix it with a consistent description beside every mention of the name, and with Organization markup that links your official profiles (https://colleviate.com/blog/organization-schema-and-sameas-for-agencies.html). ## 5. Are you answering the question that was asked? If buyers ask “how much does a website cost in India” and your pages talk about “digital experience solutions”, there is nothing to match. Write the page that answers the question in the buyer's words and gives the answer in the first paragraph; if the question is a price, lead with the range, as the buyer would want from website cost in India (https://colleviate.com/blog/website-cost-india.html). ## 6. Are you being asked in a language you have not covered? If your buyers search in Hindi, Marathi or another language, run the same questions in that language. Coverage in English does not carry over, and regional language marketing (https://colleviate.com/blog/regional-language-marketing-india.html) is a reason to check, not an afterthought. ## What not to do Do not fill pages with the phrases you want to be associated with, do not invent statistics, and do not buy mentions on pages that exist to be scraped. The click data in the SEO vs AEO (https://colleviate.com/blog/seo-vs-aeo-2026.html) debate points the same way: Pew Research Center found people clicked a traditional result in 8 percent of visits that showed an AI summary against 15 percent without one, so what you can earn is being named accurately, and that depends on being worth naming. ## Sources - GEO: Generative Engine Optimization (Aggarwal et al., KDD 2024), arXiv, June 2024: https://arxiv.org/abs/2311.09735 - Google users are less likely to click on links when an AI summary appears in the results, Pew Research Center, 22 July 2025: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/ ## Questions **Why does ChatGPT not mention my company?** Common causes: no page states plainly what you do, crawlers cannot read your pages, no independent source says it, your name collides with another, or you do not match the question asked. Test them in that order. **Does adding statistics to my pages help AI visibility?** The research paper that named the field reports visibility gains of up to 40 percent from content changes, with statistics and quotations among the strongest. Results vary by topic, so test it on your own question set. **Can I block AI crawlers by accident?** Yes. Bot rules in a CDN, firewall or security plugin can block crawlers you did not intend to block. Check them before assuming the content is the problem. --- # llms.txt: what it does, what it does not, and what to do first URL: https://colleviate.com/blog/llms-txt-what-it-does-and-what-to-do-first.html Topic: AI visibility | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** llms.txt is a proposed plain-text file that lists a site's key pages for AI systems. In June 2025 Google's John Mueller wrote that no AI system currently uses it, and we have seen no independent evidence that it changes citations. Publish one if it takes an hour, and spend the week on pages, sources and crawl access. A new file in the root of your site is the kind of fix that feels like progress. Here is what is known about this one, and why we publish it anyway. ## What it is llms.txt is a proposal for a Markdown file at the root of a website, listing the pages a language model should read first, each with a one-line description. Think of it as a curated table of contents. A longer variant, llms-full.txt, puts the text of those pages in one file. Neither is a standard that any search engine or AI company has committed to follow. ## What Google has said On 17 June 2025, John Mueller of Google wrote on Bluesky: “FWIW no AI system currently uses llms.txt.” Search Engine Roundtable reported it when the file was being added to popular plugins. Some write-ups since then say AI crawlers have or have not requested the file; we have not found a primary source that settles it either way, and several of those claims come from vendors with something to sell. ## What we do, and why We publish both files at colleviate.com/llms.txt (https://colleviate.com/llms.txt) and llms-full.txt (https://colleviate.com/llms-full.txt), generated from the same content as the site so they cannot drift. We do not claim they raise our citations, because we do not know that they do. The case for them is that they cost an hour once and nothing after, and that a clean, plain-text index of your best pages is useful to anyone building on your content, human or machine. ## What to do first instead - Make sure the pages can be read. Plain HTML text, a sensible title and one clear answer near the top, as in the technical SEO checklist (https://colleviate.com/blog/technical-seo-checklist.html). - Check crawler access. The bot rules in robots.txt, your CDN and your firewall, as in why AI answers do not mention your brand (https://colleviate.com/blog/why-ai-answers-dont-mention-your-brand.html). - Publish something worth citing. A named source, a number someone can check, original data. How AI answer engines choose sources (https://colleviate.com/blog/how-ai-answer-engines-choose-sources.html) sets out why. - Then add the file , and look in your server logs, if your host gives you them, for requests to /llms.txt. If nothing asks for it, you have your answer. ## Sources - Google says no AI system currently uses llms.txt, Search Engine Roundtable, June 2025: https://www.seroundtable.com/google-ai-llms-txt-39607.html ## Questions **Does llms.txt improve AI citations?** We have found no independent evidence that it does. In June 2025 Google's John Mueller wrote that no AI system currently uses it. It is cheap to publish, so it is a low-risk extra, not a strategy. **What is the difference between llms.txt and llms-full.txt?** llms.txt is a short list of key pages with descriptions. llms-full.txt carries the full text of those pages in a single file. Both are proposals, not standards. **Should I publish llms.txt?** If it takes an hour and stays in step with your site, yes. Do the work that has evidence behind it first: readable pages, open crawler access and citable facts. --- # Organization schema and sameAs: tell machines who you are URL: https://colleviate.com/blog/organization-schema-and-sameas-for-agencies.html Topic: AI visibility | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Organization structured data states your name, logo, URL, contact details and, in sameAs, the profiles that belong to you. Google's documentation walks through adding it to the homepage. It helps systems connect your site to your official profiles and tell you apart from similarly named businesses; it does not guarantee a knowledge panel. A name is not an identity. For a machine deciding which Colleviate you mean, the markup that links your site to your official profiles does the work a logo does for a person. ## What problem does it solve? Search engines and AI systems work with entities, not strings. If your brand name is shared with a surname, a word or another company, a system has to decide which one a page is about. Organization markup is the plain statement of the answer: this is the organisation, this is its logo, this is its website, and these are its official profiles elsewhere. Google's documentation recommends it and walks through adding it to the homepage. In November 2023 Google said it was reading more organisation details than the logo and URL it had supported since 2013, including the name, address, contact information and business identifiers. Google's own page does not promise a knowledge panel, and neither do we. ## What goes in it - name, url, logo : the same name, the canonical URL and a logo file at a stable address. - description : one plain sentence, the same one you use everywhere else. - contact : only what you publish. We list an email and a city and nothing more, which the markup permits. - sameAs : the official profiles you control. Ours lists the LinkedIn, Instagram and Facebook pages, added on 28 October 2026, and the bio on each says the same thing the site does. {"@context": "https://schema.org", "@type": "Organization", "name": "Colleviate", "url": "https://colleviate.com/", "logo": "https://colleviate.com/logo.png", "sameAs": ["https://www.linkedin.com/company/colleviate/", "https://www.instagram.com/colleviate/"]} ## Where it matters most Where trust is the product. A clinic, a nonprofit and a professional-services firm are all asked “is this real?” before they are asked anything else, and consistent identity markup is part of the answer. The ground is covered in marketing for clinics and diagnostics (https://colleviate.com/blog/marketing-for-clinics-and-diagnostics.html), nonprofit marketing in India (https://colleviate.com/blog/nonprofit-marketing-india.html) and professional services marketing (https://colleviate.com/blog/professional-services-marketing.html), and the local side in local SEO for service businesses (https://colleviate.com/blog/local-seo-for-service-businesses.html). ## Mistakes that undo it - Different names in different places. One spelling, everywhere. - sameAs pointing at profiles you do not control , or at a competitor's page that shares a word. - Markup that disagrees with the page. The footer says one thing and the schema another. - Stale profiles. A linked profile with an old description or an old address contradicts the site. Reputation is a search result (https://colleviate.com/blog/reputation-is-a-search-result.html) explains why that matters. - Treating it as a ranking trick. It is identity, not optimisation. ## Sources - Organization structured data, Google Search Central, updated September 2026: https://developers.google.com/search/docs/appearance/structured-data/organization - Expanding markup support for Organization details, including Logo structured data, Google Search Central blog, November 2023: https://developers.google.com/search/blog/2023/11/introducing-organization-markup ## Questions **What does sameAs do in Organization schema?** It lists the official profiles that belong to the organisation, such as its LinkedIn and Instagram pages, so systems can connect the website to those profiles. **Does Organization schema get me a knowledge panel?** Not by itself. Google's documentation describes the markup as supplying details about your organisation; whether a panel appears depends on far more than the markup. **Where should the markup go?** On the homepage, as Google's documentation walks through, using the same name, logo and description as the rest of your presence. --- # Internal linking for topic clusters: a practical guide and an audit URL: https://colleviate.com/blog/internal-linking-for-topic-clusters.html Topic: Method | Published: 2026-10-29 | Updated: 2026-11-03 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Internal links tell readers and crawlers which pages belong together. Build each cluster as one hub, many specific pages that each answer one question, descriptive link text from every page to the hub and to its siblings, and a visible guide on every page. Then count in-body inbound links per page and fix the ones with none first. Most content programmes publish more than they connect. The fix is a structure and a count, and the count is the part almost nobody does. ## What Google says about links Google's link guidance has two rules that matter here. Links need to be crawlable: an anchor element with an href, since that is what Google can follow. And the link text should be descriptive, in Google's words “descriptive, reasonably concise, and relevant to the page that it's on and to the page it links to.” That is the whole technical brief. Everything else is editorial judgement about which pages deserve to be joined. ## The shape of a cluster - One hub that owns the broad topic and links out to every spoke. - Spokes , each answering one question completely. If two spokes answer the same question, merge them. - Links both ways. Every spoke links to the hub and to its two or three nearest siblings, inside the text where the topic comes up, not only in a footer. - A visible guide on every page of the cluster, so a reader can see the whole set. Our share of search work is built this way. The Share of Search Index (https://colleviate.com/blog/share-of-search-index-india.html) is the hub, nine category pages and the guides are the spokes, and a guide block on all of them lists the others. Each category page links back to the hub and to the benchmark guide, and the hub links to each of them from the section that discusses it. ## The audit nobody does Crawl your own site and, for each page, count the links that point to it from inside the body text of other pages, leaving out the menu, the footer and the “related” boxes that the template adds to every page. Sort ascending. The top of that list is where your internal linking is failing. We ran it on our own Insights this month. Across 118 articles, 23 had no in-body links from any other article, a mix of sector pieces and older how-to articles, many written in a batch. They were reachable only through template boxes. That is the pattern to expect on any site that publishes in batches: the new work links to the old, and the old never learns about the new. ## How to fix a page with no inbound links - Find two articles where the topic is already mentioned and add the link there, in a sentence a reader would want. - Name the destination in the anchor, as in how to redesign a website without losing rankings (https://colleviate.com/blog/website-redesign-without-losing-rankings.html), not “click here” or “this article”. - For a market-specific page such as international SEO for Indian brands (https://colleviate.com/blog/international-seo-for-indian-brands.html), link from the pages on the process it extends, such as the technical SEO checklist (https://colleviate.com/blog/technical-seo-checklist.html). - When you publish something new, spend as long updating three old pages as you spent writing the new one. ## What to avoid Do not repeat the same exact-match anchor across dozens of pages: describe each destination as the sentence around it allows. Do not hide links in widgets that readers will never use. Do not link everything to everything; a page with sixty links tells a crawler nothing about priority. And do not count the template's “related” box as linking; it is on every page, so it is a signal about none. For earning links from other sites, which is a different job, see content that earns links (https://colleviate.com/blog/content-that-earns-links.html). Related. What happened to our own impressions as we linked and published: the first ninety days of Search Console (https://colleviate.com/blog/our-own-search-console-first-ninety-days.html). ## Sources - Link best practices for Google, Google Search Central, updated December 2025: https://developers.google.com/search/docs/crawling-indexing/links-crawlable ## Questions **What is a topic cluster?** A hub page that owns a broad topic, with spoke pages that each answer one specific question, all linked to the hub and to each other. **How do I find orphan pages?** Crawl your own site, count in-body links pointing at each page, excluding menus, footers and template "related" boxes, and sort ascending. Pages with zero are your orphans. **What makes good internal link text?** Google describes it as descriptive, reasonably concise, and relevant to both the page it is on and the page it links to. --- # Headlines that earn the click without lying: a promise-and-delivery test URL: https://colleviate.com/blog/headlines-that-earn-the-click-without-lying.html Topic: Method | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A headline earns its click when it makes one specific promise and the first screen of the page keeps it. Curiosity is legitimate and deception is not. Google's own self-assessment for content asks whether a page title gives a descriptive, helpful summary and avoids exaggerating or being shocking. A headline is a promise made in about ten words. The honest version of clickbait is a promise the first paragraph keeps. ## Where is the line? The difference is between withholding and misleading. A headline can leave out the answer, because that is what makes you open the page. It cannot promise something the page does not contain. “The number that decides whether your campaign worked” withholds. “This one trick doubles your sales” misleads. Google puts the same line in plain words. Its guidance for creating helpful content asks two questions of every page: does the main heading or page title provide a descriptive, helpful summary of the content, and does it avoid exaggerating or being shocking in nature. The practice is common, so it is worth being deliberate: in one study of the ten most-read Spanish digital newspapers, 505 of 1,321 headlines were judged clickbait. ## The test - Does the first paragraph answer what the headline raised? If the reader has to scroll to find it, the headline owed them more. - Is every number real and sourced on the page? A figure in a headline that is not in the text is the commonest lie. - Would a sceptical reader feel tricked? Read it as someone who dislikes being sold to. - Does it say who it is for? A headline that names its reader tends to attract the right readers, which is the point of a click. ## Six patterns, and their dishonest twins | Honest pattern | Example | The dishonest twin | | The specific, true number | 7.7% of revenue | A number with no source behind it | | The named tension | A 6x ROAS that lost money | An implied scandal the page does not contain | | The counterintuitive claim, with proof | The retargeting holdout | “Everything you know is wrong” | | The question the reader already has | What an agency should actually cost | A question whose answer is “buy ours” | | What we found | Nine categories, five years of data | “You will not believe what we found” | | The plain label | How to brief a performance marketing agency you can actually measure | A label that hides a sales page | The examples are titles from our own LinkedIn series and Insights, so each can be checked against the page it leads to. A number in a headline is only honest if it is in the text, as in the festive spending study (https://colleviate.com/blog/festive-spending-india-2025.html). ## Why honesty is the better commercial choice We have not seen conclusive evidence on how clickbait affects trust, so we will not claim any. Our argument is practical. A reader who arrives misled leaves, and for a business that sells considered services the reader you want is the one who stays to the end and remembers the name. For a B2B buyer the page is also a sample of the work, which is why a tone of voice that survives scrutiny (https://colleviate.com/blog/tone-of-voice-that-survives.html) and social media that sells (https://colleviate.com/blog/social-media-that-sells.html) both start from the claim, not the hook. Cultural brands that depend on curiosity, as in arts and culture marketing (https://colleviate.com/blog/arts-and-culture-marketing.html), and brands writing across languages, as in regional language marketing (https://colleviate.com/blog/regional-language-marketing-india.html), need the same test, applied in each language. ## For search results A results page shows about sixty characters of title. Put the promise in the first half, the brand in the second if it fits, and drop the brand if it does not. A title that has to be cut mid-word is a promise cut in half. ## Sources - Creating helpful, reliable, people-first content, Google Search Central, updated October 2026: https://developers.google.com/search/docs/fundamentals/creating-helpful-content - The praxis of clickbait and The Trust Project: risks and challenges for Spanish digital newspapers, Anàlisi (Flores-Vivar and Zaharía), 2022: https://www.doi.org/10.5565/rev/analisi.3463 ## Questions **What is ethical clickbait?** A headline that creates real curiosity and then keeps its promise: it withholds the answer but does not misrepresent the page. The test is whether the first paragraph delivers what the headline raised. **What does Google say about page titles?** Its guidance for creating helpful content asks whether the main heading or page title provides a descriptive, helpful summary of the content, and whether it avoids exaggerating or being shocking. **How long should a title be for search results?** About sixty characters is what a results page typically shows, so put the promise first and the brand last. --- # How B2B buyers shortlist agencies before they contact one URL: https://colleviate.com/blog/how-b2b-buyers-shortlist-agencies.html Topic: Method | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** 6sense's 2025 study of B2B buyers found first contact with a vendor happens about 61 percent of the way through the buying process, that buying groups fill most of their shortlist on day one, and, in the company's data, that the winning vendor was on that day-one list 95 percent of the time. For an agency, discovery has to work before anyone contacts you. By the time a prospect writes to you, most of the decision has been made without you in the room. The work is to be in the room earlier, without being there. ## What the study found 6sense's 2025 B2B Buyer Experience Report, as summarised by CustomerThink, found that the average first contact between buyer and seller now happens when the buyer is 61 percent of the way through the buying process, down from 69 percent in earlier studies. Buying groups evaluated about five vendors and filled four of those five places on their shortlist on the first day. 94 percent of buyers ranked their shortlist by preference before they contacted anyone, and 6sense's own page puts the winner's presence on the day-one list at 95 percent. Read it with care. The data is 6sense's, drawn from its own buyer survey and platform, in markets where it sells, mainly software and technology services. It is not an Indian agency benchmark, and the company itself describes the ratios as directional. The direction, though, is consistent with how considered purchases work: people form a view first and use the first conversation to confirm it. ## What that means for an agency If the shortlist is mostly set before contact, then a prospect who writes to you is already leaning your way, and a prospect who never writes to you may have ruled you out without you knowing. Your marketing is judged in the silent part of the process. Three things build the shortlist: - Prior experience and recommendation. People shortlist agencies they know or were told about. This is the oldest channel and the strongest, and it is why the first ninety days with an agency (https://colleviate.com/blog/first-ninety-days-with-an-agency.html) and how an engagement ends (https://colleviate.com/blog/when-to-fire-your-agency.html) are marketing too. - Search and AI answers. The question “who should we hire for this” is now put to assistants as well as search engines. Whether you are named is measurable, as in how to track whether AI answers name your brand (https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html). - Things a buyer can inspect without talking to you. Published prices or ranges, as in a pricing page that sells (https://colleviate.com/blog/pricing-page-that-sells.html), and work you can describe in detail. ## Where it matters most In sectors where the buyer is a specialist and the purchase is considered, as in manufacturing and industrial marketing (https://colleviate.com/blog/manufacturing-and-industrial-marketing.html) and professional services marketing (https://colleviate.com/blog/professional-services-marketing.html), the shortlist is built from a small number of trusted names and a few deliberate searches. A brand that exists only in paid media is easy to skip. Share of search for B2B (https://colleviate.com/blog/share-of-search-for-b2b.html) is one way to see whether yours is in the set at all. ## A test you can run - Ask your last five clients, in a sentence, how they first heard of you and who else they considered. You will learn more from five answers than from a funnel report. - Ask the same of the last three you lost. - Run your category's buyer questions through the main assistants and search, ten times each, and see whether you appear. - Write down the first place a stranger could verify you. If it is a contact form, there is nothing yet for them to inspect. None of this replaces the rest of the programme, as B2B lead generation in India (https://colleviate.com/blog/b2b-lead-generation-india.html) sets out. It changes what the programme is trying to do: be known well enough before the first call that the first call is short. ## Sources - Research round-up: 6sense study provides critical insights on B2B buyer behaviour (secondary report), CustomerThink, 2025: https://customerthink.com/research-round-up-6sense-study-provides-critical-insights-on-b2b-buyer-behavior/ ## Questions **When do B2B buyers first contact a vendor?** 6sense's 2025 study found first contact happens about 61 percent of the way through the buying process, down from 69 percent in earlier studies. The data is 6sense's own and not specific to India or to agencies. **How many vendors make the shortlist?** In the same study, buying groups evaluated about five vendors and filled four of the five places on their shortlist on the first day. **What should an agency do about it?** Be findable and credible before anyone contacts you: in recommendations, in search and AI answers, and in material a buyer can inspect without a call. --- # What eBay learned when it switched off its brand ads URL: https://colleviate.com/blog/what-ebay-learned-by-switching-off-brand-ads.html Topic: Measurement | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** In a field experiment published in Econometrica in 2015, eBay stopped paying for search ads on queries containing its own name, and the sales moved to the free listing. The authors found no measurable short-term benefit from the brand ads. The lesson is not to switch ads off, but to test which of your ads are doing work before you believe the platform report. Most of the money spent on brand keywords buys clicks you were going to get anyway. eBay checked, and then published the result. ## What did eBay switch off? Three researchers, Thomas Blake, Chris Nosko and Steven Tadelis, asked eBay a plain question: what happens to sales if the search ads stop? They answered it by stopping them. The paper, Consumer Heterogeneity and Paid Search Effectiveness , appeared in Econometrica in January 2015. The first result is the one worth remembering. When eBay stopped paying for ads on searches that contained its own name, the sales didn't go away. People clicked the free listing instead. The authors report no measurable short-term benefit from the brand ads. Put plainly, the ads were being credited with sales that would have happened anyway. The ad platform's report would have said otherwise, because a report counts a purchase that follows a click as a purchase the ad made. That is the gap the ROAS reality check (https://colleviate.com/blog/performance-marketing-roas-reality.html) describes, shown at the scale of one of the largest advertisers online. ## What about the other keywords? The team also paused bidding on non-brand keywords, the generic searches, for 60 days across a 30 percent sample of eBay's US traffic. They report that the measured returns were a fraction of what a standard, non-experimental estimate suggested. The reason is plain once you see it: people who search for a product are, more often than not, people who were already inclined to buy it. One version of the abstract adds that new and infrequent users were positively influenced by the ads. Read that as a clue about where ads do work, not as a rule. ## What does it mean for you? Concede the obvious first. eBay is one of the best known names on the internet, and a brand nobody has heard of will see something different. The authors say as much: ads reach people who don't know you yet. So this isn't a case for turning your ads off. It's a case for knowing which of them are working. The useful split is the one in branded and category search (https://colleviate.com/blog/branded-versus-category-search.html). Brand searches are people who already know you. Category searches are people who don't. They deserve different budgets and different tests. ## Run the test yourself - Pick two cities that behave alike , with similar sales and similar search volume for your brand. - Pause brand ads in one of them for four to six weeks and leave the other alone. - Judge by total conversions from brand searches, paid and free together. The platform's own number will always say the ads worked. - Check who else bids on your name. If a rival buys it, their ad may sit above your free listing, and pausing could cost you sales. That's our reasoning, not a finding of the paper, which is why you test it. - Move the saving to something that reaches people who don't know you yet, and measure that the same way. The mechanics of the account are in Google Ads account structure (https://colleviate.com/blog/google-ads-account-structure.html), and the same logic applies to retargeting (https://colleviate.com/blog/retargeting-limits.html). The paper's value isn't the result. It's the habit: switch something off, in a small place, for a short time, and look at what really happens. ## Sources - Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment (Blake, Nosko and Tadelis), Econometrica 83(1), January 2015: https://www.econometricsociety.org/publications/econometrica/2015/01/01/consumer-heterogeneity-and-paid-search-effectiveness-large - Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment (working paper w20171), National Bureau of Economic Research, 2014: https://www.nber.org/papers/w20171 ## Questions **Do brand keyword ads work?** In eBay's field experiment, stopping ads on searches containing its own name produced no measurable short-term loss, because sales moved to the free listing. A lesser known brand may see different results, so test your own. **How do I test whether my brand ads are incremental?** Pause them in one of two similar cities for four to six weeks and compare total conversions from brand searches, paid and free combined, not the ad platform's own figure. **Does this mean paid search does not work?** No. The authors report returns from non-brand keywords were a fraction of non-experimental estimates, and that new and infrequent users were positively influenced by ads. The point is to measure, not to stop. --- # Does advertising build awareness, or belief? What the panel data says URL: https://colleviate.com/blog/does-advertising-build-awareness-or-belief.html Topic: Measurement | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A panel study of more than 300 brands, by Clark, Doraszelski and Draganska, found that advertising had a consistently significant positive effect on brand awareness but no significant effect on perceived quality. Treat advertising as the job of being known, and give the job of being believed to proof: product, reviews, case studies and independent mentions. Advertising buys familiarity. Whether people believe you is decided somewhere else, and a brief that confuses the two spends money on the wrong one. ## What did the study measure? Clark, Doraszelski and Draganska built a panel of more than three hundred brands. For each one they had annual advertising spend, and survey measures of brand awareness and perceived quality from a large consumer survey. They treated advertising as an investment in two separate stocks, one of awareness and one of perceived quality, and asked how much each moved when spend changed. Because they followed the same brands over time, they could set aside the fixed differences between brands, and they used dynamic panel methods to deal with the fact that firms choose their ad spend for reasons that aren't random. Their finding is short. Advertising had a consistently significant positive effect on brand awareness. It had no significant effect on perceived quality. ## What that does and doesn't say It's one study, on annual data, using survey measures, across the brands and categories in that survey. It doesn't prove that no advertisement has ever changed what anyone thinks of a product. It's a fair account of what the money does on average: it makes more people know you exist. That's more useful than it sounds, because it's the cleaner of the two jobs to measure and to buy. Awareness can be bought with reach. Belief can't. ## Split the brief in two Awareness without proof produces visits that don't convert. Proof without awareness produces a library nobody opens. Most budgets fund one and hope for the other. The practical answer is to give each job its own owner, its own measure and its own money. - The awareness job. Reach, frequency and the share of people in the category who know you. A cheap, public way to watch it is share of search (https://colleviate.com/blog/share-of-search-leading-indicator.html), which is our own proxy for awareness, not part of the study. - The belief job. Reviews you didn't write, case studies with real numbers, independent coverage, and a product that does what the page says. PR worth paying for (https://colleviate.com/blog/pr-worth-paying-for.html) and reputation as a search result (https://colleviate.com/blog/reputation-is-a-search-result.html) are about earning exactly this. ## What to change on Monday Look at the last campaign brief. If it says “build trust” and the only deliverable is reach, the brief has asked for something the money doesn't buy. Rewrite it so reach is the awareness target, and add a separate line, with a different budget, for what proves the claim. For the long-run balance between the two, see the brand and performance split (https://colleviate.com/blog/brand-versus-performance-split.html). ## Sources - The Effect of Advertising on Brand Awareness and Perceived Quality: An Empirical Investigation using Panel Data (Clark, Doraszelski and Draganska), Quantitative Marketing and Economics, 2009: https://repository.upenn.edu/marketing_papers/303 ## Questions **Does advertising change perceived quality?** In a panel study of more than 300 brands by Clark, Doraszelski and Draganska, advertising had no significant effect on perceived quality, though it had a consistently significant positive effect on awareness. It is one study, so treat it as the average, not a law. **What builds perceived quality, then?** The study does not say. Our reading is proof you did not write: reviews, independent coverage, case studies with real numbers and a product that does what the page says. **How do I measure brand awareness cheaply?** Share of branded search against your named competitors is a free, public proxy. It is our proxy, not part of the study, and it measures attention rather than belief. --- # How blended retention can rise while the business gets worse URL: https://colleviate.com/blog/how-blended-retention-rises-while-the-business-gets-worse.html Topic: Measurement | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Blended retention is an average across customers of different ages. If you acquire fewer new customers, the base fills with older ones, who leave less often, so the average rises even when every group is doing the same or worse. Report retention by signup cohort, not as one blended figure. The most flattering number on a subscription dashboard is often a symptom. Here is the sum that shows why, and the chart that stops it fooling the board. This article is arithmetic, not a study. The numbers are invented to make the mechanism visible. They are not from any client. ## The set-up Take a subscription business with two kinds of customer. Customers in their first year leave often: say 30 percent of them are gone by the end of the year, so 70 percent are retained. Customers who have been there longer leave less: say 92 percent are retained. Year one, while acquisition is strong. The base is 1,000 older customers and 1,000 first-year customers. The older group keeps 920. The newer group keeps 700. That's 1,620 of 2,000, a blended retention of 81 percent. Year two, after acquisition slows. Marketing has had a bad year, and only 300 new customers arrived. The base is now 1,400 older customers and 300 first-year customers. The older group keeps 1,288. The newer group keeps 210. That's 1,498 of 1,700, a blended retention of 88 percent. The dashboard says retention rose seven points. Neither group got any better. The business simply stopped adding the customers who leave. ## Make it worse Now let the first-year customers get worse as well, because the last campaign pulled in poorer-fit buyers. Say only 60 percent of them stay. Year two becomes 1,288 plus 180, which is 1,468 of 1,700, or 86 percent. So blended retention is still five points above where it started, while new-customer retention fell ten points and new-customer volume fell by seventy percent. The headline is up. The business is shrinking. In other words, the number is answering a question nobody asked: what share of the customers we happen to have are staying? ## The same trap in usage and revenue The same mix effect hides in any ratio of a mixed population. Daily over monthly active users rises when casual users leave the monthly count faster than heavy users leave the daily one. Net revenue retention rises when a few large, loyal accounts dominate the base. Whenever a dashboard shows one blended rate for customers of different ages or sizes, ask what the mix was last year. ## The table that stops it Report retention by the month or quarter customers signed up, and read down the columns. Each row is a cohort, each column is how long they've been with you. If the newest rows sit below the old ones at the same age, you have a problem the blend is hiding. If you can only add one thing to the monthly pack, add this, and put it beside the acquisition count, because the two together are the story. The wider habit is in a marketing dashboard that gets read (https://colleviate.com/blog/marketing-dashboard-that-gets-read.html) and a reporting cadence (https://colleviate.com/blog/marketing-reporting-cadence.html) the board will keep. It matters most in subscription and software businesses, as in SaaS marketing in India (https://colleviate.com/blog/saas-marketing-india.html), and in any D2C brand living on repeat orders, as in D2C growth after the first crore (https://colleviate.com/blog/d2c-growth-after-the-first-crore.html). And it is one more reason data-driven marketing (https://colleviate.com/blog/data-driven-marketing-definition.html) has to mean a decision ranked by evidence, not a metric that happens to look good. The wider gap between what is spent and what can be proved is the subject of the marketing measurement gap (https://colleviate.com/blog/marketing-measurement-gap.html). ## Questions **Why does blended retention rise when acquisition slows?** Because the base fills with older customers, who leave less often than new ones. The average improves even if no group has improved. **What should I report instead of blended retention?** Retention by signup cohort, shown beside the number of customers acquired in each period. **Does the same problem affect DAU/MAU and net revenue retention?** It can. Any ratio over a mixed population shifts when the mix shifts, so ask what the mix was last year. --- # Measuring a business that sells through dealers, distributors and referrers URL: https://colleviate.com/blog/measuring-a-business-that-sells-through-others.html Topic: Growth audits | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** If most of your sales run through dealers, distributors, advisers or referring doctors, your website analytics describe the smallest channel and the partners' effort goes unmeasured. Measure the partner layer directly: concentration of sales among partners, partner attrition, how enquiries that reach you leak to rivals, who recommends you and why, and what each partner can actually see of you. A dashboard measures what it can see. In a business that sells through other people, what it can see is usually the smallest part. This is our reasoning from audit work, not a study, and it cites no figures. We'd rather say so than dress it up. ## The dashboard problem A leadership team looks at what it can see. For most companies that's web traffic, ad spend and leads. Now take a business that sells through dealers, a mutual fund house that sells through advisers, an equipment maker with distributors, or a hospital that depends on referring doctors. In those businesses the sale is made by someone else, in a conversation the company never sees. So the sharpest numbers describe the smallest channel, and the biggest channel runs on belief. The dashboard is accurate. The picture is wrong. Both are true at once, which is what makes it hard to argue with. ## What goes unmanaged When effort is made by a partner, the company can't observe how hard that partner tried. It sees the order, or the lack of one. It can't tell a partner who never mentions you from one who mentions you and loses. A distributor with ten brands on the shelf gives each of them whatever attention it earns, and nobody is writing that down. ## Five measures for the part you can't see - Concentration. What share of sales comes from your top ten partners? If it's most of them, you have ten relationships, not a market. - Attrition. How many partners that sold in one year sold nothing the next? A quiet exodus doesn't show in a national total until it's late. - Leakage. Of the enquiries you receive directly, how many go to a partner, and how many end with a rival? Track the handoff, since that's where interest is lost. - Recommendation. Ask partners, in a short call, what they say about you and to whom. You will learn more from twelve of those than from a survey of end customers. - Visibility. What can a partner find, in five minutes, to help them sell you? Pricing, specifications, proof, a name to call. If it's thin, they sell something else. ## Where it shows up The pattern is strongest where trust is passed along a chain. Manufacturing and industrial marketing (https://colleviate.com/blog/manufacturing-and-industrial-marketing.html) runs through distributors and specifiers. Automotive marketing (https://colleviate.com/blog/automotive-marketing-india.html) runs through dealers. Clinics and diagnostics depend on referrers, as in marketing for clinics and diagnostics (https://colleviate.com/blog/marketing-for-clinics-and-diagnostics.html), and the same is true of healthcare marketing (https://colleviate.com/blog/healthcare-marketing-india.html) more widely. Financial products sold through advisers share it, as fintech marketing in India (https://colleviate.com/blog/fintech-marketing-india.html) describes. How the person making the recommendation forms a shortlist is in how B2B buyers shortlist (https://colleviate.com/blog/how-b2b-buyers-shortlist-agencies.html). None of this makes website analytics useless. It makes them one voice in a larger conversation. For the demand side of the same problem, see B2B lead generation in India (https://colleviate.com/blog/b2b-lead-generation-india.html). ## Questions **How do you measure marketing when most sales go through partners?** Measure the partner layer directly: concentration of sales among partners, partner attrition, leakage of enquiries to rivals, what partners say about you, and what a partner can find to help them sell you. **Why is website analytics a poor guide in a dealer or distributor business?** It describes the channel the company controls, which is usually the smallest, and says nothing about the effort partners make or fail to make. **What is partner concentration?** The share of sales that comes from your largest partners. When it is high, a few relationships carry the business, and losing one is a bigger event than any campaign result. --- # Iconic Indian ads for the ages: a cultural history, 1966 to the 1990s URL: https://colleviate.com/blog/iconic-indian-ads-for-the-ages.html Topic: Ads for the ages | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Indian advertising between 1966 and 1990 moved from the hoarding to the television set to the national film, and each medium changed what an ad could be: a daily public commentary, a sponsor inside a story, a state message in commercial form. Scholars of the culture industries, from Adorno and Horkheimer to Mazzarella and Mankekar, give us the tools to read them. Advertising is the part of culture that pays for the rest of it. These are five moments when Indian advertising did something larger than selling, and what the scholarship says about them. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. They are homages to the idioms of their decades, drawn for this series, and each is labelled as an original. ## Why read old advertising at all? Because it is the most honest archive of what a country was told it wanted. Two traditions help. Adorno and Horkheimer, writing in 1944, argued that mass-produced culture works by standardisation: products and entertainments that differ on the surface and agree underneath. They called the surface difference pseudo-individuality. Raymond Williams, writing in New Left Review in 1960, called advertising a magic system, one that dresses ordinary goods in meanings and, in his view, obscures the real choices behind them. Neither was writing about India, and both can be argued with. Used carefully, they give us questions to ask of an old ad: what is it standardising, and what is it dressing up? ## The five moments - 1966, the hoarding. A butter brand turns the outdoor billboard into a daily joke about the news. Read the essay (https://colleviate.com/blog/amul-hoardings-and-the-topical-ad.html). - 1969 onwards, the price. A detergent built for households the market had ignored, and a jingle that carried it. Read the essay (https://colleviate.com/blog/nirma-and-the-price-of-a-jingle.html). - 1984, the sponsor inside the story. A serial on the state broadcaster carries a brand's name. Read the essay (https://colleviate.com/blog/hum-log-and-the-sponsored-serial.html). - 1988, the state as advertiser. A national film in the form of a commercial. Read the essay (https://colleviate.com/blog/mile-sur-mera-tumhara-and-the-state-as-advertiser.html). - The 1980s and 1990s, the agency. Who decided what “Indian” looked like in an ad. Read the essay (https://colleviate.com/blog/who-writes-indianness-the-ad-agency-as-culture-industry.html). And a method, so you can do this yourself: how to read a vintage ad in five questions (https://colleviate.com/blog/how-to-read-a-vintage-ad.html). ## A medium map | Decade | The medium that mattered | What it let an ad do | | 1960s | Outdoor hoardings and print | Be seen daily, and be cheap enough to change weekly | | 1970s | Print, and the first television commercials | Show the product in a life, not just in a pack | | 1980s | Doordarshan, in colour from 1982 | Reach a whole household at once, and attach a name to a story | | 1990s | Satellite and cable | Fight for attention among many channels instead of one | ## Seven originals For each decade we drew a period piece in its printing idiom, with caricature figures, for an invented brand. They are on Instagram and LinkedIn as a series, and each is credited as an original. If one of them stirs a memory of a hoarding, a serial or a scooter ride, that is the purpose of the genre, and the reason these ads outlived their products. ## A note on our sources The history is drawn from trade press and publisher descriptions, and where they disagree on a date or a credit we say so. The scholarship is cited as its publishers and reviewers describe it. Where something is our reading, we label it. If you can correct a detail, please write to us. ## Sources - The Culture Industry: Enlightenment as Mass Deception (Adorno and Horkheimer), Marxists Internet Archive, 1944 text: https://www.marxists.org/reference/archive/adorno/1944/culture-industry.htm - Advertising: The Magic System (Raymond Williams), New Left Review, 1960: https://newleftreview.org/issues/I4/articles/raymond-williams-the-magic-system - Shoveling Smoke: Advertising and Globalization in Contemporary India (William Mazzarella), Duke University Press, 2003: https://www.dukeupress.edu/shoveling-smoke - Screening Culture, Viewing Politics: An Ethnography of Television, Womanhood, and Nation in Postcolonial India (Purnima Mankekar), Duke University Press, 1999: https://www.dukeupress.edu/screening-culture-viewing-politics - Doordarshan, Wikipedia, undated: https://en.wikipedia.org/wiki/Doordarshan ## Questions **What is the culture industry?** A term from Adorno and Horkheimer's 1944 essay for the mass production of culture, which they argued standardises products and entertainments while presenting small surface differences as individuality. **Which Indian ads does this series cover?** The Amul topical hoardings from 1966, Nirma from 1969, the sponsored serial Hum Log in 1984, the 1988 film Mile Sur Mera Tumhara, and the agencies that shaped how advertising spoke about India. **Are the artworks copies of real ads?** No. They are original period-style pieces for fictional brands, drawn for this series. --- # The hoarding that answered back: Amul and the topical ad URL: https://colleviate.com/blog/amul-hoardings-and-the-topical-ad.html Topic: Ads for the ages | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Amul's hoarding campaign began in 1966, created by Sylvester daCunha's agency ASP, and became known for topical ads that comment on current events beside the butter. It is widely reported as the longest-running advertising campaign in the world. Read through the culture industry critics, it is advertising that shows its working. Most advertising asks for attention. This campaign earned it by commenting on the day, which is a different and harder thing. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The piece beside this essay is a 1960s-style hoarding for a fictional butter. ## What happened in 1966 According to the brand's own account and trade-press retellings, Sylvester daCunha's agency, ASP, won the Amul butter account in 1966. Television and print were costly, so the agency chose outdoor hoardings. The art director, Eustace Fernandes, and daCunha built a mascot meant to appeal to mothers and children. The first topical hoarding, reported as “Thoroughbread”, showed her as a jockey in the racing season. The tagline is credited to the daCunha family, with Nisha daCunha's “Utterly Amul” extended by Sylvester to the form we know. Sources disagree on whether the first hoarding appeared in 1966 or 1967, and on exactly how the account reached the agency, so we say “1966” because most of them do. The campaign is described by the brand and in the press as the longest-running advertising campaign in the world, and as holding a Guinness record. We haven't seen the record's own listing, so we pass that on as reported. ## What a topical ad does An ordinary ad repeats. A topical ad is a different animal: it changes as the country does. Each hoarding takes something the street is already talking about and puts the butter beside it. The ad stops being a message from a company and becomes a column in a daily paper that happens to be a billboard. Somebody in 1966 got to write a pun for a living and call it strategy, and the rest of us have been jealous ever since. ## Two critics, two readings Adorno and Horkheimer argued that the culture industry offers pseudo-individuality: small surface differences on a standardised product. The topical hoarding seems an obvious case. The product and the mascot never change, and only the pun does. Raymond Williams's magic system points the same way: the butter is dressed in the glow of whatever the news happened to be. Now concede what they would miss. The joke is public, signed and open to argument. It doesn't hide that it is an ad. It uses the form of an editorial to say the one thing an advertiser has to say, which is that it is paying attention. That is our reading, not the critics' own. It is also why the form lasted: a standardised product with a visibly human voice is a harder thing to resent. ## What it asks of a brand today Three things. A mascot with a point of view that doesn't change. A cost structure that lets you produce on the day, which in 1966 meant cheap outdoor. And a willingness to be public. Most brands have the first and fail the other two, which is the argument in what a campaign owes the culture it borrows from (https://colleviate.com/blog/what-a-campaign-owes-the-culture.html). For where the idea comes from, see the brief is not the idea (https://colleviate.com/blog/the-brief-is-not-the-idea.html). ## Sources - Amul topical story, Amul, undated: https://www.amul.com/m/amul-topical-story - Revisiting Sylvester daCunha's Utterly Butterly years with Amul, afaqs, undated: https://www.afaqs.com/news/advertising/revisiting-sylvester-dacunhas-utterly-butterly-years-with-amul - Sylvester da Cunha, Wikipedia, undated: https://en.wikipedia.org/wiki/Sylvester_da_Cunha - The Culture Industry: Enlightenment as Mass Deception (Adorno and Horkheimer), Marxists Internet Archive, 1944 text: https://www.marxists.org/reference/archive/adorno/1944/culture-industry.htm - Advertising: The Magic System (Raymond Williams), New Left Review, 1960: https://newleftreview.org/issues/I4/articles/raymond-williams-the-magic-system ## Questions **When did the Amul topical ads begin?** Most sources date the first hoarding to 1966, created by Sylvester daCunha's agency ASP, though some give 1967. **Why did Amul use hoardings?** Accounts say television and print were too costly for the account, so the agency chose the more affordable outdoor medium. **Is it the longest-running ad campaign?** It is widely reported as the longest-running advertising campaign in the world, with a Guinness record. We have not seen the record listing itself. --- # When the sponsor joined the story: Hum Log and Doordarshan URL: https://colleviate.com/blog/hum-log-and-the-sponsored-serial.html Topic: Ads for the ages | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Hum Log began on Doordarshan on 7 July 1984 and is widely reported as the first Indian show to carry sponsorship, with Nestle among the sponsors. It marks the point where advertising stopped being an interval between programmes and became part of how a programme was made and paid for. A single channel, a single national audience and a brand that paid to sit beside the story. The 1980s were the decade advertising got a seat in the living room. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The 1980s piece beside this essay is a fictional sponsored serial card. ## What the record says Doordarshan began carrying commercials in the 1970s, with sources giving 1976 or 1977. Colour arrived in August 1982, around the Asian Games in Delhi, and the Games drove a wave of set purchases. Hum Log, a family serial, began on 7 July 1984 and is widely reported as the first Indian show to carry sponsorship. A Nestle advertising executive, interviewed by Storyboard18, recalled that sponsored programmes did not exist at the time, that the company took up the sponsorship after the producer approached it, and that the sponsor's advertising time after each episode ran to about 40 seconds. Nestle's Maggi launch film had already run on Doordarshan in 1983. Two cautions. The “first sponsored show” claim rests mainly on a Nestle recollection and encyclopaedia entries, and sources disagree on when sponsored programmes became routine (1984 or 1987). And the serial itself was reportedly not well preserved, so there is little left to watch. ## A sponsor is not an interval Before sponsorship, an advertiser bought time between things. After it, an advertiser bought a relationship with a thing. A family serial watched by a household at a fixed hour gave a brand the same slot every week, and the audience came to associate the two. The product became a regular guest, the sort who turns up every Tuesday at the same time and is somehow always fed. ## What the scholarship adds Purnima Mankekar's Screening Culture, Viewing Politics is an ethnography of how women viewers in Delhi received Doordarshan serials of the 1980s and early 1990s. Its argument, as its publisher and reviewers describe it, is that television helped build ideas of nation and womanhood, and that the state invested in an image of the woman to define the nation. It is a study of the serials, not of the advertising around them, and we are extending it, not quoting it. Our extension is this. If the serial taught a household what a good family looked like, the sponsor was the voice that told it what to buy to be one. Raymond Williams's point that advertising obscures the choices behind goods has a sharp edge here, because the story's authority passed to the product. ## What it means now Every branded series, creator partnership and podcast is a descendant of this arrangement. The question is the old one: is the brand a guest in the story, or has it taken over the telling? Our view is the first, as what a campaign owes the culture it borrows from (https://colleviate.com/blog/what-a-campaign-owes-the-culture.html) argues, and as a brand launch plan (https://colleviate.com/blog/brand-launch-campaign-plan.html) should say in writing. ## Sources - How Nestle sponsored Hum Log and changed the TV advertising landscape in India, Storyboard18, undated: https://www.storyboard18.com/quantum-brief/how-nestle-sponsored-hum-log-and-changed-the-tv-advertising-landscape-in-india-7275.htm - 1982 Asian Games: how sports brought colours to Indian television, The Bridge, undated: https://thebridge.in/news/1982-asian-games-how-sports-brought-colours-to-indian-television/ - Doordarshan, Wikipedia, undated: https://en.wikipedia.org/wiki/Doordarshan - Screening Culture, Viewing Politics: An Ethnography of Television, Womanhood, and Nation in Postcolonial India (Purnima Mankekar), Duke University Press, 1999: https://www.dukeupress.edu/screening-culture-viewing-politics - Advertising: The Magic System (Raymond Williams), New Left Review, 1960: https://newleftreview.org/issues/I4/articles/raymond-williams-the-magic-system ## Questions **Was Hum Log the first sponsored show in India?** It is widely reported as such, with Nestle among its sponsors, on the evidence of a Nestle executive's recollection and reference entries. Sources differ on when sponsored programmes became routine. **When did Doordarshan start carrying ads?** In the 1970s; sources give 1976 or 1977. **What did Mankekar study?** Her 1999 book is an ethnography of how women viewers in Delhi received Doordarshan serials of the 1980s and early 1990s, and how television helped build ideas of nation and womanhood. It does not study the ads. --- # The price of a jingle: Nirma and the market nobody served URL: https://colleviate.com/blog/nirma-and-the-price-of-a-jingle.html Topic: Ads for the ages | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Nirma began in 1969 when Karsanbhai Patel started making a detergent in Ahmedabad and priced it at a fraction of the market leader. Reported accounts say it went on to take a large share of India's detergent market and forced Hindustan Lever into a new product. The lesson is about price and address: whom an ad is for. Most advertising of the 1970s spoke to households that could already afford the brand. One speaking to households that couldn't found a very large audience. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The piece beside this essay is a 1970s-style soap advertisement for a fictional brand. ## What the accounts report Karsanbhai Patel, a laboratory technician in Ahmedabad, began making a phosphate-free detergent in 1969 in a room at his home and sold it door to door. Accounts put his launch price at about Rs 3 to 3.5 a kilo, against about Rs 13 to 15 for the leading brand, Surf. They agree that it grew very quickly, and several say it reached about 60 percent of the detergent market by 1988, a figure that appears without a clear primary source, so treat it as reported. Hindustan Lever's answer is reported to have been the Wheel detergent in 1988, in an effort internally named Operation STING. A fuller account is a case study by Ghoshal and Butler, which we have not read. A widely repeated anecdote that Hindustan Lever's research had spotted the opportunity and ignored it is, in the source we found, a hypothesis and not a documented fact. We leave it out. ## Who an ad is for Advertising of the period, as the culture industry critics would have predicted, tended to address a household already in the market: a woman choosing between brands of similar standing. Nirma's price addressed a different household, one for whom the choice was between a detergent and a bar of soap. The brand's real innovation was its audience, and nobody had put it in a brief. The advertising's job was to make a cheap thing feel like a chosen thing, which is where a memorable jingle and a repeated slogan did the work a premium pack couldn't. ## Standardisation, from below Adorno and Horkheimer worried that mass culture flattened differences. The Nirma story shows a mass product being used the other way: it widened who could take part in the market for a branded good. William Mazzarella's study of the Bombay advertising industry in the 1980s and 1990s is, as its publisher describes it, about how advertising mediated mass consumerism and national identity. We read Nirma as an early case of that mediation reaching further down the income ladder than the agencies of the time usually aimed. That is our reading, not his. ## What it means now If your price is your idea, the ad should say so plainly and often. If your category has ignored a segment, the creative problem is the segment, not the category. And a larger rival's reaction takes years, which is how a small brand gets its head start. For the economics, see the brand and performance split (https://colleviate.com/blog/brand-versus-performance-split.html), and for how a price ought to be explained, what a marketing agency should cost (https://colleviate.com/blog/what-a-marketing-agency-should-cost.html). ## Sources - How a brilliant Gujarati created 'Sabki Pasand Nirma' in his backyard, The Better India, undated: https://www.thebetterindia.com/171504/gujarat-nirma-karsanbhai-patel-make-in-india/ - Roots to legacy: why is 'Sabki Pasand Nirma', Impact, undated: https://www.impactonnet.com/impact-stories/roots-to-legacy-why-is-sabki-pasand-nirma-16495.html - The Culture Industry: Enlightenment as Mass Deception (Adorno and Horkheimer), Marxists Internet Archive, 1944 text: https://www.marxists.org/reference/archive/adorno/1944/culture-industry.htm - Shoveling Smoke: Advertising and Globalization in Contemporary India (William Mazzarella), Duke University Press, 2003: https://www.dukeupress.edu/shoveling-smoke ## Questions **When did Nirma start?** In 1969, when Karsanbhai Patel began making a phosphate-free detergent in Ahmedabad. **How cheap was it?** Accounts vary, but put its launch price at roughly Rs 3 to 3.5 a kilo against about Rs 13 to 15 for the leading brand. **What was Hindustan Lever's response?** Reported accounts say it launched Wheel in 1988 as part of an effort called Operation STING. A Ghoshal and Butler case study is the more rigorous source; we have not read it. --- # One tune, many voices: Mile Sur Mera Tumhara and the state as advertiser URL: https://colleviate.com/blog/mile-sur-mera-tumhara-and-the-state-as-advertiser.html Topic: Ads for the ages | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Mile Sur Mera Tumhara was first aired on Doordarshan on 15 August 1988, after the Prime Minister's Independence Day speech, developed by Lok Seva Sanchar Parishad as a national integration message. Credits for the song and the lyrics differ between sources. It used the grammar of advertising, repetition, a tune and a face, to sell an idea. A film in the form of a commercial, broadcast to a nation, with no product in it. It is the most remembered piece of Indian advertising craft that sold nothing. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The piece beside this essay is a fictional 1970s family-scooter advertisement, chosen because its idea, everyone together, is the one this film carries. ## What the record says The film was developed in 1988 by Lok Seva Sanchar Parishad and promoted by Doordarshan and the Ministry of Information. It was first aired on 15 August 1988, after the telecast of the Prime Minister's speech from the Red Fort. Credits are contested. Suresh Mullick is widely credited with conceiving the song, and some sources credit Piyush Pandey with the lyrics and Kailash Surendranath with direction, while sources disagree on the composer. A 2025 report on a later attempt to remake it concluded the original was hard to repeat. We don't resolve the credits here, and because they differ, we won't name the lyricist or composer as fact. ## What it borrowed from advertising Repetition. A refrain that returns in every language. A parade of faces that makes the argument by accumulation: if enough different people sing the same line, the line is true. These are the techniques of a commercial, applied to an abstraction. The state borrowed the market's grammar to sell an idea no market could, which is a compliment to the grammar. ## The culture industry reading Adorno and Horkheimer would see a standardised form carrying a manufactured sameness. We think that is only half right. The film's form is standard, but its content is plurality: many languages, many places, one melody. The uniformity is in the structure, and the variety is in the surface, which is the reverse of the pseudo-individuality they described. Purnima Mankekar's work on 1980s Doordarshan, as publishers and reviewers describe it, is about how television helped build a national identity that sat uneasily with the differences among its viewers. This film is the most optimistic version of that project. Reading it as advertising, not just as anthem, shows how much of nation-building was a craft problem: how do you make a million different people feel they are in one picture? ## What it means for brands Brands that claim to speak to all of India have a harder version of the same problem. The honest ones do what the film did: they put the differences on screen and find the line that holds them. The dishonest ones translate one ad into twelve languages and call it inclusive. For the practical side, see regional language marketing in India (https://colleviate.com/blog/regional-language-marketing-india.html) and what a campaign owes the culture it borrows from (https://colleviate.com/blog/what-a-campaign-owes-the-culture.html). ## Sources - Ogilvy celebrates two decades of 'Mile Sur Mera Tumhara', exchange4media, undated: https://www.exchange4media.com/advertising-news/ogilvy-celebrates-two-decades-of-mile-sur-mera-tumhara-32162.html - Mile Sur Mera Tumhara, Wikipedia, undated: https://en.wikipedia.org/wiki/Mile_Sur_Mera_Tumhara - The Culture Industry: Enlightenment as Mass Deception (Adorno and Horkheimer), Marxists Internet Archive, 1944 text: https://www.marxists.org/reference/archive/adorno/1944/culture-industry.htm - Screening Culture, Viewing Politics: An Ethnography of Television, Womanhood, and Nation in Postcolonial India (Purnima Mankekar), Duke University Press, 1999: https://www.dukeupress.edu/screening-culture-viewing-politics ## Questions **When was Mile Sur Mera Tumhara first broadcast?** On 15 August 1988 on Doordarshan, after the Prime Minister's Independence Day speech. **Who made it?** It was developed by Lok Seva Sanchar Parishad and promoted by Doordarshan and the Ministry of Information. Credits for the song, lyrics and direction differ between sources. **Why read it as advertising?** Because it used the grammar of a commercial, repetition, a refrain and a parade of faces, to carry an idea of national unity rather than a product. --- # Who writes Indianness? The ad agency as culture industry URL: https://colleviate.com/blog/who-writes-indianness-the-ad-agency-as-culture-industry.html Topic: Ads for the ages | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** William Mazzarella's 2003 book Shoveling Smoke is an ethnography of the Bombay advertising business and its role in the growth of mass consumerism in 1980s and 1990s India. As its publisher and reviewers describe it, it treats agencies as mediators between global brands and local culture, and shows why that mediation never reduces to a formula. When a brand says it is speaking for India, someone in an agency had to decide what India sounds like. That someone is the subject of one of the best books on Indian advertising. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The 1990s piece beside this essay is a fictional cable-television advertisement. This essay relies on how the publisher and reviewers describe the book, not on a chapter-by-chapter reading, and says only what they agree on. ## The book William Mazzarella's Shoveling Smoke: Advertising and Globalization in Contemporary India , published by Duke University Press in 2003, is an ethnography of the Bombay advertising business. It looks at how the industry took part in the rise of mass consumerism in the 1980s and 1990s, and at several campaigns in a major Bombay agency and what they say about Indian national identity. Reviewers describe it as a critical anthropology of the culture industries and as an argument against treating globalisation as one totalising force, because a stubborn local excess keeps interrupting it. ## Why the agency is the interesting place An agency sits between a company that wants to sell the same thing everywhere and a public that does not live everywhere in the same way. Someone has to decide what “Indian” looks like in the ad. That decision is made by a small group of people in a few cities, under time pressure, for a client. A great deal of what a country sees of itself in commercials has passed through a meeting room, usually one with bad coffee and a deadline. ## The culture industry, reconsidered Adorno and Horkheimer pictured the culture industry as a smooth machine. The ethnographic view, as we understand it from the reviews, is rougher: a machine made of people who argue, guess and disagree, and who are also members of the audience they address. That is why the output keeps surprising the system that produced it. It is also why the 1990s are interesting. Satellite and cable multiplied channels, and with them the number of rooms in which that decision was being made. ## What a brand should ask its agency - Who decided what India looks like in this ad, and whom does that exclude? - Where, in the film or the page, does the local get to speak in its own terms and not as a flavour? - If we translated this idea, what would we lose, and have we shown it on screen? The craft questions sit beside the film that tried to answer them at national scale, Mile Sur Mera Tumhara (https://colleviate.com/blog/mile-sur-mera-tumhara-and-the-state-as-advertiser.html), and the working questions in regional language marketing in India (https://colleviate.com/blog/regional-language-marketing-india.html). ## Sources - Shoveling Smoke: Advertising and Globalization in Contemporary India (William Mazzarella), Duke University Press, 2003: https://www.dukeupress.edu/shoveling-smoke - The Culture Industry: Enlightenment as Mass Deception (Adorno and Horkheimer), Marxists Internet Archive, 1944 text: https://www.marxists.org/reference/archive/adorno/1944/culture-industry.htm - Screening Culture, Viewing Politics: An Ethnography of Television, Womanhood, and Nation in Postcolonial India (Purnima Mankekar), Duke University Press, 1999: https://www.dukeupress.edu/screening-culture-viewing-politics ## Questions **What is Shoveling Smoke about?** As its publisher and reviewers describe it, an ethnography of the Bombay advertising business and its role in mass consumerism in 1980s and 1990s India, and what its campaigns say about national identity. **What does Mazzarella argue about globalisation?** Reviewers describe the book as challenging a totalising account of globalisation, because a local excess keeps interrupting it. **Why does the agency matter?** It sits between global brands and local culture, so a great deal of what a country sees of itself in commercials passes through the choices of a few people there. --- # How to read a vintage ad in five questions URL: https://colleviate.com/blog/how-to-read-a-vintage-ad.html Topic: Ads for the ages | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Ask five things of any old ad: who is addressed, what is promised, what is left out, what the medium allowed, and what the ad assumes about the person looking at it. The method draws on Raymond Williams's idea of advertising as a magic system and on Adorno and Horkheimer's account of standardisation. An old ad is a document. These five questions turn it from something to smile at into something to learn from. ## The five questions - Who is addressed? Not who is in the picture, but who the ad assumes is looking. Age, income, language and the room the ad expects to be read in. - What is promised? One thing, said as plainly as you can. If it takes two sentences, the ad promised two things and one of them is the real one. - What is left out? Williams argued that advertising obscures real choices. So ask what choice is hidden: the price, the alternative, the cost. - What did the medium allow? A hoarding is read in three seconds from a moving vehicle, a print ad in thirty, a television spot in a household. The form is half the message. - What does it take for granted? The best evidence of a period is what the ad didn't need to explain. ## A worked example, on our own fiction Take the 1980s piece in this series, a fictional noodle brand that promises to be “ready before the ad break ends”. - Who: a mother, watching television, in a household for whom a break is a unit of time. - Promise: speed, said in the language of the viewer's own evening. - Left out: cost, nutrition, and the fact that someone still has to cook it. - Medium: a print ad borrowing the rhythm of television, which tells you television had become the clock of the house. - Taken for granted: that the break exists, is short, and is shared. In other words, the ad's most interesting claim is about the viewer's life, not the product. That is the reading Adorno and Horkheimer's critique invites and Williams's refines: look past the product to the world the ad assumes. ## Using it on your own work The same five questions applied to a current campaign are a cheap audit. They are the first page of an outside-in brand audit (https://colleviate.com/blog/outside-in-brand-audit.html), and they explain why the brief is not the idea (https://colleviate.com/blog/the-brief-is-not-the-idea.html). The cultural history behind them is in the series hub, iconic Indian ads for the ages (https://colleviate.com/blog/iconic-indian-ads-for-the-ages.html), and one of the clearest examples is the Amul hoardings (https://colleviate.com/blog/amul-hoardings-and-the-topical-ad.html). ## Sources - Advertising: The Magic System (Raymond Williams), New Left Review, 1960: https://newleftreview.org/issues/I4/articles/raymond-williams-the-magic-system - The Culture Industry: Enlightenment as Mass Deception (Adorno and Horkheimer), Marxists Internet Archive, 1944 text: https://www.marxists.org/reference/archive/adorno/1944/culture-industry.htm ## Questions **How do you analyse an old advertisement?** Ask five questions: who is addressed, what is promised, what is left out, what the medium allowed, and what the ad takes for granted. **What is the magic system?** Raymond Williams's name for modern advertising, which he described as a highly organised system of magical inducements and satisfactions that obscures real choices. **What is pseudo-individuality?** Adorno and Horkheimer's term for small surface differences that make standardised products look individual. --- # Retro, rebooted: five new-age categories, five retro campaigns, and a vote URL: https://colleviate.com/blog/retro-rebooted-window-shopping-for-new-age-brands.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Retro, rebooted is a series of five campaign concepts that take a vintage Indian advertising idea and redraw it for today, one each for an EV maker, a payments app, an investing app, a quick-commerce app and a food delivery app. We name no client. We ask people to tag the brand they think should run each, and we publish the tally. Our own Share of Search Index supplies the category data. The old ads in our Time Machine series worked because they were simple, human and local. New-age brands have the same needs and a different look. We drew the idea; you tell us whose it is. This is part of #ColleviateTimeMachine. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The five pieces below are different: they are drawn for today, in a clean contemporary style, for categories of company. No real logo, mascot, pack or app screen appears. Brands are named in the captions only. ## How the vote works Each concept is posted on Instagram and LinkedIn with one question: which brand should run this? Tag them. We count the tags for each concept and publish the tally here, and we'll update this page. If a brand wants to talk, it can write to us, and a real version would be developed with that brand, not for it. ## 1. An EV maker: “Sab ek saath. Now with no phatphat.” The retro idiom is the 1970s family scooter, the most recognisable image on an Indian road. The idea keeps the whole family on one vehicle and takes away the noise, so silence becomes the thing the household shares. The data: in our index of electric two-wheeler search, Ather holds 62.2 percent of branded search, ahead of Ola Electric (14.7), TVS iQube (13.5) and Bajaj Chetak (9.5), and has led every month since September 2025. See the electric two-wheeler deep-dive (https://colleviate.com/blog/share-of-search-electric-two-wheelers-india.html). ## 2. A payments app: “Khata closed. Scan open.” The retro idiom is the kirana shop's credit ledger. The idea sets the closed ledger beside a QR board in the same uncle's hand. It speaks to the shop, not the shopper. The data: in UPI app search, PhonePe holds 37.0 percent and Paytm 35.6, with Google Pay at 21.5 and CRED at 5.9, and PhonePe has led since May 2026, per the UPI deep-dive (https://colleviate.com/blog/share-of-search-upi-payments-india.html). A close race is where an idea like this earns its keep. ## 3. An investing app: “Start small. Stay long.” The retro idiom is the bank passbook and the coin jar. The idea makes beginning feel ordinary, with an uncle, a passbook and a plant in a jar. The data: Groww holds 40.1 percent of stockbroking search, ahead of Zerodha (30.2) and Angel One (21.2), per the stockbroking deep-dive (https://colleviate.com/blog/share-of-search-stockbroking-india.html). A real campaign would need to follow the regulator's rules for investment advertising, and nothing here is advice. ## 4. A quick-commerce app: “Faster than your chai.” The retro idiom is the neighbourhood kirana that knows your order. The idea puts a ten-minute stopwatch beside a cup of chai and the same uncle with a bag. The data: Blinkit holds 56.4 percent of quick-commerce search, Zepto 28.0, Swiggy Instamart 10.7 and BigBasket 4.9, per the quick-commerce deep-dive (https://colleviate.com/blog/share-of-search-quick-commerce-india.html). ## 5. A food delivery app: “Amma packed it. Maps tracked it.” The retro idiom is the home tiffin. The idea sets a mother packing a tiffin beside a phone with a live route. The data: Swiggy holds 50.9 percent and Zomato 49.1 of company-brand search, and the lead has changed hands repeatedly, per the food delivery deep-dive (https://colleviate.com/blog/share-of-search-food-delivery-india.html). It is the closest race in the index, and the one where a warm idea could matter most. ## What the data does and doesn't say Share of search measures attention, not sales, and it is a reading of a brand name against its named rivals. It doesn't tell a brand what to say. It tells us which of these categories is a close race, which is a settled one, and so what kind of idea each might want. How to read a number like this is in what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html). ## How we would test any of them Before spending a rupee, we would run each concept against one question: does it make the brand more likely to be searched for by name? We'd measure that before and after, against the same competitors, with the margin of error shown, as in how to measure share of search (https://colleviate.com/blog/how-to-measure-share-of-search.html). The reading method behind the format is in how to read a vintage ad (https://colleviate.com/blog/how-to-read-a-vintage-ad.html), and the history is in the series hub, iconic Indian ads for the ages (https://colleviate.com/blog/iconic-indian-ads-for-the-ages.html). ## The open door If you work at one of these brands and this is useful, or wrong, we'd like to hear it: colleviate.com/contact (https://colleviate.com/contact). ## Sources - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What is Retro, rebooted?** A series of five campaign concepts that redraw a vintage Indian advertising idea for today, one each for an EV maker, a payments app, an investing app, a quick-commerce app and a food delivery app. **Do the brands know about this?** No. These are original concepts and no brand has seen them. A tag is a suggestion from the public. **Where does the data come from?** From the Colleviate Share of Search Index for India, which uses Google Trends. It measures attention, not sales. --- # Window shopping: the brands we would love to work with, and the campaigns we would pitch URL: https://colleviate.com/blog/window-shopping.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Window shopping is a Colleviate series in which we choose a brand we admire, read the idea buried in its heritage, and draw one campaign we would pitch for today. Five concepts are out so far, for an EV maker, a biscuit, an investing app, a paint brand and a hotel group. You can nominate a brand, tag someone who works at one, or tell us the idea is wrong. Every agency has a list of brands it wishes were clients. Ours is public. Here is how it works, what we have drawn so far, and how you can help a good idea find its owner. This is part of #ColleviateTimeMachine. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. ## The premise Most agency work starts when a client calls. Window shopping starts earlier. We look at a brand we would like to work with, find the idea in its own history that still has something to say, and draw how it might look today. It is a way to show our thinking in public, and to find out whether the thinking holds up when strangers read it. ## How each concept is built - A heritage idea. One thing the brand did in the past that people still remember, read from sources we link to. - A reading. What that idea was really doing, in plain terms. We separate what the sources report from what we think. - Today's version. One campaign idea and one picture, drawn from scratch. - A number. Where the brand stands in search, from our own Share of Search Index (https://colleviate.com/blog/share-of-search-index-india.html), because an idea for a close race is different from an idea for a settled one. ## The five concepts so far | Brand | Heritage idea | Our concept | Read more | | Ather | The family scooter and its famous slogan | “Now with no phatphat”: silence as the thing a household shares | From Hamara Bajaj to the electric scooter (https://colleviate.com/blog/hamara-bajaj-to-the-electric-scooter.html) | | Parle | The glucose biscuit and the tea cup | “One more dunk”: the ritual, not the biscuit | Parle-G and the chai ritual (https://colleviate.com/blog/parle-g-and-the-chai-ritual.html) | | Groww | The passbook and the coin jar | “Start small, stay long”: beginning made ordinary | Savings as a story (https://colleviate.com/blog/savings-as-a-story-before-groww.html) | | Birla Opus | A paint mascot who stayed a boy for decades | “A new coat for an old wall”: the transformation shot | Gattu and the transformation ad (https://colleviate.com/blog/gattu-and-the-transformation-ad.html) | | Taj Hotels | The view as the product | “Book the view”: the window as the pitch | The wider Retro, rebooted set (https://colleviate.com/blog/retro-rebooted-window-shopping-for-new-age-brands.html) | Two more essays sit under the series because they explain the method: how a soap borrowed the film star (https://colleviate.com/blog/lux-and-the-borrowed-glamour-of-the-film-star.html), and why nostalgia is not a strategy (https://colleviate.com/blog/nostalgia-is-not-a-strategy.html). The vintage history behind the format is in iconic Indian ads for the ages (https://colleviate.com/blog/iconic-indian-ads-for-the-ages.html), and the five fictional-brand concepts for new categories are in Retro, rebooted (https://colleviate.com/blog/retro-rebooted-window-shopping-for-new-age-brands.html). ## How you can take part - Nominate. Tell us a brand whose heritage deserves this treatment. - Tag. If you know someone at one of these brands, tag them on the post. A tag is the whole mechanism. - Connect. If you are that person, say so. We would like to hear what is right and what is wrong. - Debate. If the idea is weak, say why. A better argument beats a polite like. - Vote. On each post we ask which of two directions you prefer, and we publish the result. ## What this is not It is not a claim about any brand's strategy, and the numbers measure attention, not sales. A real version would be developed with the brand, not for it. Colleviate is window shopping. The right brief could make it real. To start a conversation, write to colleviate.com/contact (https://colleviate.com/contact). ## Sources - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What is Window shopping?** A Colleviate series in which we choose a brand we admire, read the idea in its heritage and draw the campaign we would pitch for today. **How do I take part?** Nominate a brand, tag someone you know who works at one of the brands, join the debate on the posts, or vote between two directions. **Where does the data come from?** From the Colleviate Share of Search Index for India, which uses Google Trends. It measures attention, not sales. --- # The soap that borrowed the film star: celebrity as proof URL: https://colleviate.com/blog/lux-and-the-borrowed-glamour-of-the-film-star.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Lux soap was introduced in India in 1929 and became known for putting film actresses in its advertising. The idea is borrowed proof: the brand does not say the soap is good, it lets a face people already admire stand in for the claim. Sources disagree on which actress came first and when. A star in an ad is a claim without an argument. It works because we already trust the star. That is worth understanding before you pay for one. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The soap on the page is a fictional one. ## What the sources report Lux was introduced in India in 1929, according to the brand's encyclopedia entry, and its Indian advertising is widely described as an early and lasting use of film actresses as endorsers. The actress Leela Chitnis is often named as the first Indian film star to appear in its advertising, in the early 1940s. Accounts give different years, and we have not found a primary record that settles it, so we say only that she is reported among the first. ## How borrowed proof works A soap cannot easily prove it is better. The buyer cannot compare it before buying, and the difference, if any, is small. A star solves the problem sideways. The ad does not argue. It lends the product the attention and trust the star already has. In our reading this is the plainest case of what Raymond Williams called the magic system: the product is dressed in a meaning that has nothing to do with its chemistry. Adorno and Horkheimer would add that the pattern is standardised. Swap the face and the structure stays, which is why the form lasted. We offer that as a reading of the format, not as a judgement of any person who appeared in it. ## Where it breaks Borrowed proof is only as good as the borrowing. It fails when the star is seen to endorse everything, when the audience stops believing the arrangement, or when the glamour belongs more to the star than to the product, so that people remember the face and not the brand. The test is simple: cover the logo, and ask whether anyone could say what was being sold. ## What a brand can do with it today Look for proof that is earned, not borrowed. A real customer, a real result, or a ritual the product belongs in all carry more weight than a famous face. If you do use a face, make the product the thing the face is doing, not standing beside. The wider habit of reading an ad for what it dresses up is in how to read a vintage ad in five questions (https://colleviate.com/blog/how-to-read-a-vintage-ad.html), and the agencies behind these choices are in who writes Indianness? (https://colleviate.com/blog/who-writes-indianness-the-ad-agency-as-culture-industry.html). The series that this belongs to is Window shopping (https://colleviate.com/blog/window-shopping.html). ## Sources - Lux (soap), Wikipedia, undated: https://en.wikipedia.org/wiki/Lux_(soap) - Leela Chitnis, Wikipedia, undated: https://en.wikipedia.org/wiki/Leela_Chitnis - The Culture Industry: Enlightenment as Mass Deception (Adorno and Horkheimer), Marxists Internet Archive, 1944 text: https://www.marxists.org/reference/archive/adorno/1944/culture-industry.htm - Advertising: The Magic System (Raymond Williams), New Left Review, 1960: https://newleftreview.org/issues/I4/articles/raymond-williams-the-magic-system ## Questions **When did Lux arrive in India?** Its encyclopedia entry gives 1929. **Who was the first film star in Lux advertising in India?** Leela Chitnis is often named, in the early 1940s, but accounts differ on the year and we have not seen a primary record. **Does celebrity endorsement work?** It can lend attention and trust to a product, but it fails when the face is remembered and the brand is not. --- # Parle-G and the chai ritual: when the product is the occasion URL: https://colleviate.com/blog/parle-g-and-the-chai-ritual.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Parle was founded in 1929 and its glucose biscuit was launched in 1939, later renamed Parle-G. The brand has explained the G as glucose and, in later years, as genius. The idea we would draw is the ritual: the dunk, the tea and the pause, not the biscuit alone. The strongest brands sell an occasion, not a product. The biscuit is the excuse for the pause. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The concept on the page is ours, not Parle's. ## What the record says Parle Products was founded in 1929 in Vile Parle, Mumbai. The glucose biscuit was launched in 1939 as Parle Gluco and was later renamed Parle-G. Sources differ on when the renaming happened, giving the early 1980s, and we have not seen a primary record, so we leave the year open. The G stood first for glucose. Later campaigns have also read it as genius, and we pass that on as the brand's own later gloss, not as its origin. ## What the brand sells Strip the packaging and the claim is modest: a plain, cheap biscuit. What turned it into a habit is the setting it sits in. It belongs with tea, with a certain hour, with a certain kind of waiting. When a product becomes the occasion, the competition changes. It stops competing with other biscuits and starts competing with the pause itself. ## The idea we would draw “One more dunk.” A grandfather, a rainy window, a cup, and a child learning the correct time to lift the biscuit. The picture carries the ritual, and the handwritten note on it carries the joke that everyone has lost one to the chai. No claim is made about the biscuit. The ritual does the work. ## What it asks of a brand - Find the moment the product already lives in, and name it without explaining it. - Let the humour belong to the user, not the brand. - Keep the idea cheap to repeat. A ritual works by repetition. ## The caution Ritual is not the same as nostalgia. A ritual is practised today. Nostalgia is remembered. We take that distinction up in nostalgia is not a strategy (https://colleviate.com/blog/nostalgia-is-not-a-strategy.html), and for how a brand can sit inside the calendar, see what a brand owes a festival (https://colleviate.com/blog/what-a-brand-owes-a-festival.html). If you know someone at Parle, tag them on the post. More concepts are on the Window shopping page (https://colleviate.com/blog/window-shopping.html). ## Sources - Parle-G, Wikipedia, undated: https://en.wikipedia.org/wiki/Parle-G - Parle Products, Wikipedia, undated: https://en.wikipedia.org/wiki/Parle_Products ## Questions **When was Parle-G launched?** As Parle Gluco in 1939, according to its encyclopedia entry. The renaming date is given differently by different sources. **What does the G stand for?** Originally glucose. Later campaigns have also read it as genius. **Is this a Parle campaign?** No. It is a concept drawn by Colleviate for the Window shopping series. --- # Savings as a story: the passbook, the coin jar and the first investment URL: https://colleviate.com/blog/savings-as-a-story-before-groww.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** In our Share of Search Index, Groww holds 40.1 percent of stockbroking search, ahead of Zerodha at 30.2 and Angel One at 21.2. The idea we would draw for a beginner investing brand is not a promise of returns but a picture of beginning: a passbook, a jar and a plant. This is a concept, and not financial advice. The hardest customer for an investing app is not the expert. It is the person who thinks the whole thing is for someone else. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. Nothing here is investment advice, and a real campaign would need to follow the regulator's rules for investment advertising. ## The category today In our stockbroking deep-dive (https://colleviate.com/blog/share-of-search-stockbroking-india.html), Groww holds 40.1 percent of branded search, with Zerodha at 30.2 and Angel One at 21.2. Share of search measures attention, not sales, and a lead of ten points is a reading of a moment, not a verdict. It does tell us that the category is contested and that being remembered by name is already a battle. ## The story the old household told For a generation the first act of saving was physical: a passbook with the bank's stamp, a jar on a shelf, a relative who said to start small. It was slow, visible and unglamorous. It also made beginning feel ordinary, which is the very thing a new investor finds hardest. ## The idea we would draw “Start small. Stay long.” An uncle holds a passbook beside a jar in which a plant grows. A handwritten note on the card has a parent telling a child to save, as parents do. The picture does not promise a return. It makes the first step look like something a family already does. ## The test Would a first-time investor feel less alone after seeing it? That is the question the idea must pass, and it is cheaper to ask people than to guess. Before any spend, we would measure whether the idea makes the brand more likely to be searched by name, against the same rivals, as set out in how to measure share of search (https://colleviate.com/blog/how-to-measure-share-of-search.html). ## What it asks of a brand - Say less about returns and more about beginning. - Use the language of the household, not the exchange. - Keep the regulator's disclaimers where they belong, and do not hide them in the art. If you know someone at Groww, tag them on the post. More concepts are on the Window shopping page (https://colleviate.com/blog/window-shopping.html). ## Sources - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What share of stockbroking search does Groww hold?** In the Colleviate Share of Search Index, 40.1 percent, ahead of Zerodha at 30.2 and Angel One at 21.2. **Is this investment advice?** No. It is a creative concept and a reading of search data. Share of search measures attention, not performance. **Is this a Groww campaign?** No. It is a concept drawn by Colleviate for the Window shopping series. --- # Gattu and the transformation ad: why a painted wall makes a good story URL: https://colleviate.com/blog/gattu-and-the-transformation-ad.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Asian Paints' mascot Gattu is reported to have been created by the cartoonist R. K. Laxman in 1954 and retired in 2002. Paint is a category of transformation, and in our Share of Search Index the incumbent holds 72.1 percent of search against 11.0 for the newer Birla Opus. The concept we would draw is a transformation shot. Paint is the only category where the product becomes the room. That makes the before and after the strongest picture it owns. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The painter on the page is our own character, not the mascot named here. ## What the record says Asian Paints is widely reported to have introduced the character Gattu in 1954, drawn by the cartoonist R. K. Laxman, and to have retired him in 2002. We have taken those dates from reference entries and haven't seen the company's own archive, so treat them as reported. The more interesting fact is the one the mascot illustrates: a paint brand was for decades a character with a face, not a colour card. ## The category today In our paints deep-dive (https://colleviate.com/blog/share-of-search-paints-india.html), the incumbent holds 72.1 percent of branded search. Birla Opus is at 11.0, Nerolac at 7.6 and Berger at 7.3. A challenger at 11 percent is not trying to out-shout a leader at 72. It is trying to be remembered for one thing. Share of search measures attention, not sales. ## The idea we would draw “A new coat for an old wall.” A painter on a ladder rolls bands of colour onto a wall that has seen better years. A handwritten note on the card says the thing every household says, which is that the painting will be done after the festival. Paint's most honest competitor is not another brand. It is the decision to wait. ## What a transformation ad asks of a brand - Show the change, not the can. - Give the audience a person to watch, so they have someone to root for. - Address the delay. The real objection is rarely the price. The topical, public voice that Amul used in the same era is in the hoarding that answered back (https://colleviate.com/blog/amul-hoardings-and-the-topical-ad.html). If you know someone at Birla Opus, tag them on the post. More concepts are on the Window shopping page (https://colleviate.com/blog/window-shopping.html). ## Sources - Asian Paints, Wikipedia, undated: https://en.wikipedia.org/wiki/Asian_Paints - R. K. Laxman, Wikipedia, undated: https://en.wikipedia.org/wiki/R._K._Laxman - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **Who created Gattu?** The cartoonist R. K. Laxman is reported to have created the Asian Paints mascot in 1954. We have not seen the company archive. **What share of paints search does Birla Opus hold?** In the Colleviate Share of Search Index, 11.0 percent, against 72.1 for Asian Paints, 7.6 for Nerolac and 7.3 for Berger. **Is this a Birla Opus campaign?** No. It is a concept drawn by Colleviate for the Window shopping series. --- # From Hamara Bajaj to the electric scooter: what a family vehicle used to promise URL: https://colleviate.com/blog/hamara-bajaj-to-the-electric-scooter.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** The Hamara Bajaj campaign for the Bajaj Chetak scooter is dated to 1989 and is reported as created by the agency Lintas. In our Share of Search Index, Ather holds 62.2 percent of electric two-wheeler search. The concept we would draw is a family on one scooter, with silence as the thing they share. The scooter ad that people still hum was not about the scooter. It was about the people on it. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. The family on the page is ours, on an orange scooter that is nobody's. ## What the record says The Hamara Bajaj campaign is dated to 1989 and is associated with the Bajaj Chetak scooter. It is reported as the work of the agency Lintas, and trade press credits particular people on the creative side. We haven't verified individual credits, so we do not name anyone here. What is not in dispute is how long the line lasted in memory. It said “ours”, and a whole country knew who was meant. ## What it was selling It was not selling a scooter's specification. It was selling belonging: the sense that this was the vehicle of a certain kind of household, the one that goes to the same relatives and the same market. That is a stronger claim than speed, and it explains why the memory outlasted the product's place in the market. It is also the point at which nostalgia becomes a trap, which we take up in nostalgia is not a strategy (https://colleviate.com/blog/nostalgia-is-not-a-strategy.html). ## The category today In our electric two-wheeler deep-dive (https://colleviate.com/blog/share-of-search-electric-two-wheelers-india.html), Ather holds 62.2 percent of branded search, ahead of Ola Electric at 14.7, TVS iQube at 13.5 and Bajaj Chetak at 9.5. Share of search measures attention, not sales. A clear lead is a different problem from a close race: the job is to keep what you hold and make it mean something. ## The idea we would draw “Now with no phatphat.” A family of four on one scooter, riding in silence, with “shhh” on its side. The handwritten note on the card says what every family knows, which is that whoever sits in the middle is the luggage. The old promise was belonging. The new one is that the household gets to share quiet. ## What it asks of a brand - Sell the household, not the specification. - Make the new feature (here, silence) the thing the family shares. - Be funny about the real inconvenience. The luggage joke is why the card works. If you know someone at Ather, tag them on the post. More concepts are on the Window shopping page (https://colleviate.com/blog/window-shopping.html). ## Sources - Hamara Bajaj, Wikipedia, undated: https://en.wikipedia.org/wiki/Hamara_Bajaj - Bajaj Chetak, Wikipedia, undated: https://en.wikipedia.org/wiki/Bajaj_Chetak - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **When did Hamara Bajaj begin?** It is dated to 1989 and is associated with the Bajaj Chetak scooter. **What share of electric two-wheeler search does Ather hold?** In the Colleviate Share of Search Index, 62.2 percent, ahead of Ola Electric at 14.7, TVS iQube at 13.5 and Bajaj Chetak at 9.5. **Is this an Ather campaign?** No. It is a concept drawn by Colleviate for the Window shopping series. --- # Nostalgia is not a strategy: what retro advertising can and cannot do URL: https://colleviate.com/blog/nostalgia-is-not-a-strategy.html Topic: Window shopping | Published: 2026-10-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A 2014 study in the Journal of Consumer Research by Lasaleta, Sedikides and Vohs reports that nostalgia weakened the desire for money in their experiments, and that nostalgic participants were more willing to spend. That makes nostalgia a useful mood, not a strategy. A strategy still needs an audience, a claim and a measure. Making people feel something warm about the past is easy. Making that warmth do a commercial job is the part that needs thought. The artwork in this series is original and shows fictional brands. We have not redrawn any real campaign, mascot or slogan, and no national symbol appears. ## What the research reports Lasaleta, Sedikides and Vohs published “Nostalgia Weakens the Desire for Money” in the Journal of Consumer Research in 2014. As the authors describe it, across a set of experiments, people who were made to feel nostalgic placed less value on money and were more willing to spend. These were laboratory studies with specific tasks and participants, so the finding does not tell us what a retro campaign will do for a given brand in a given market. ## What that does and doesn't support It supports a modest claim: warm memories can loosen the grip of price in the moment. It does not support the much larger claims that sometimes get made for retro campaigns, that they build brands or convert buyers by themselves. A feeling is a condition for persuasion, not the persuasion. ## Three questions before you go retro - Whose memory is it? A look that stirs your senior team may mean nothing to the customer you want. In our vintage series (https://colleviate.com/blog/iconic-indian-ads-for-the-ages.html), the people who recognise a 1980s serial are not the people who open an investing app for the first time. - What is the claim? The retro look is a wrapper. Underneath there must be a reason to choose the brand today. If you cannot say it plainly with the wrapper removed, you do not have a campaign. - How will you know? Name the number before you start. For a brand idea, that can be whether people look for it by name more than before, against the same rivals. ## The cultural caution Borrowing from the past also means borrowing its assumptions. William Mazzarella's study of Indian advertising reminds us that an ad is a negotiation about who the audience is. A retro treatment inherits the old answer unless you change it. That is why our Window shopping (https://colleviate.com/blog/window-shopping.html) concepts draw new characters instead of redrawing old ones, and why the rest of the habit is in how to read a vintage ad (https://colleviate.com/blog/how-to-read-a-vintage-ad.html). ## Where it works Retro works best when the idea underneath is still true: a ritual still practised, a worry still felt, a household still gathered. It works worst as decoration. For the longer argument about what a campaign owes the culture it borrows from, see what a campaign owes the culture (https://colleviate.com/blog/what-a-campaign-owes-the-culture.html). ## Sources - Nostalgia Weakens the Desire for Money (Lasaleta, Sedikides and Vohs), Journal of Consumer Research 41(3), 713-729, University of Chicago Press, 2014: https://doi.org/10.1086/677227 - Shoveling Smoke: Advertising and Globalization in Contemporary India (William Mazzarella), Duke University Press, 2003: https://www.dukeupress.edu/shoveling-smoke - Advertising: The Magic System (Raymond Williams), New Left Review, 1960: https://newleftreview.org/issues/I4/articles/raymond-williams-the-magic-system ## Questions **Does nostalgia make people spend more?** In the 2014 experiments by Lasaleta, Sedikides and Vohs, nostalgic participants valued money less and were more willing to spend. They were laboratory studies, so the result does not transfer automatically to a campaign. **Is retro advertising a strategy?** No. It is a style. A strategy needs an audience, a claim and a measure. **When does retro work?** When the idea underneath is still true today, such as a ritual still practised. --- # Blinkit vs Zepto vs Swiggy Instamart: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-quick-commerce-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Blinkit has the largest share of quick-commerce brand search in India: 56.4% over the three months to September 2026, against Zepto at 28.0%. It has led every month since February 2025. Over the past year Blinkit gained most, +5.6 points, while Zepto lost most, -3.6. Quick commerce went from an experiment to a daily habit inside five years, and branded search shows which name India came to think of first. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/quick-commerce.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads quick commerce search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | Blinkit | 13.1% | 50.8% | 56.4% | +5.6 pts | | Zepto | 21.1% | 31.6% | 28.0% | -3.6 pts | | Swiggy Instamart | 21.4% | 12.3% | 10.7% | -1.6 pts | | BigBasket | 44.5% | 5.2% | 4.9% | -0.3 pts | BlinkitZeptoSwiggy InstamartBigBasket Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed BigBasket held 44.5% of the set's branded search when the series begins and 4.9% now. Blinkit rose from 13.1% to 56.4% (the brand took the Blinkit name in December 2021, so its line starts from zero) and has led every month since February 2025, after Zepto held the lead from October 2024 to January 2025. Zepto sits at 28.0% and Swiggy Instamart at 10.7%. Read BigBasket with care: it is a scheduled grocery service as well as a quick-commerce one, so part of its fall is likely the category changing shape rather than a brand losing. ## Who has led, and for how long The lead has changed hands 5 times in 59 months. Counting every month since November 2021: Blinkit for 44 months, BigBasket for 11 months and Zepto for 4 months. | Leader | From | To | Months | Public event that month | | BigBasket | Nov 2021 | May 2022 | 7 | | | Blinkit | Jun 2022 | Jun 2022 | 1 | | | BigBasket | Jul 2022 | Oct 2022 | 4 | | | Blinkit | Nov 2022 | Sep 2024 | 23 | | | Zepto | Oct 2024 | Jan 2025 | 4 | | | Blinkit | Feb 2025 | Sep 2026 | 20 | | ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - December 2021: BigBasket -20.5 points, to 41.0%, the month Grofers became Blinkit and the new name began drawing searches. - February 2024: Blinkit +12.1 points, to 60.9%. - January 2022: BigBasket -10.0 points, to 31.0%. Each brand's single highest week of search interest: Blinkit in August 2026, Zepto in August 2026, Swiggy Instamart in February 2026 and BigBasket in August 2026. ## Search share and market share Eternal, Blinkit's parent, and Swiggy, which owns Instamart, report quick-commerce gross order value in their quarterly results. Read a widening search lead against those figures, and against Zepto's own disclosures where it makes them, rather than against press estimates of market share, which vary with the metric (orders, gross merchandise value or gross order value) and with the analyst. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “Blinkit”, “Zepto”, “Instamart” and “BigBasket”. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | Blinkit | Zepto | Swiggy Instamart | BigBasket | | Sep 2026 | 56.4 | 28.1 | 10.7 | 4.8 | | Aug 2026 | 55.7 | 27.9 | 11.1 | 5.3 | | Jul 2026 | 57.2 | 28.0 | 10.2 | 4.7 | | Jun 2026 | 52.1 | 31.4 | 10.7 | 5.8 | | May 2026 | 56.0 | 29.1 | 9.9 | 5.0 | | Apr 2026 | 52.6 | 30.5 | 11.5 | 5.4 | | Mar 2026 | 51.9 | 30.1 | 12.2 | 5.8 | | Feb 2026 | 51.5 | 30.1 | 13.3 | 5.1 | | Jan 2026 | 52.9 | 28.0 | 13.0 | 6.1 | | Dec 2025 | 52.5 | 28.4 | 13.2 | 5.9 | | Nov 2025 | 50.4 | 29.9 | 13.7 | 6.1 | | Oct 2025 | 51.5 | 30.1 | 13.1 | 5.4 | | Sep 2025 | 50.7 | 30.9 | 13.3 | 5.1 | | Aug 2025 | 50.2 | 32.6 | 12.1 | 5.1 | | Jul 2025 | 51.6 | 31.3 | 11.5 | 5.5 | | Jun 2025 | 49.8 | 32.7 | 11.8 | 5.7 | | May 2025 | 47.6 | 35.7 | 11.3 | 5.5 | | Apr 2025 | 46.1 | 36.8 | 11.8 | 5.3 | | Mar 2025 | 42.6 | 41.2 | 11.2 | 5.0 | | Feb 2025 | 42.1 | 41.8 | 11.6 | 4.5 | | Jan 2025 | 40.1 | 45.0 | 10.4 | 4.6 | | Dec 2024 | 39.9 | 45.3 | 9.7 | 5.1 | | Nov 2024 | 39.8 | 44.4 | 8.9 | 6.9 | | Oct 2024 | 41.7 | 42.1 | 9.6 | 6.6 | | Sep 2024 | 47.9 | 33.3 | 10.6 | 8.2 | | Aug 2024 | 48.2 | 32.4 | 10.4 | 9.0 | | Jul 2024 | 46.5 | 33.0 | 9.7 | 10.8 | | Jun 2024 | 50.5 | 27.6 | 9.7 | 12.2 | | May 2024 | 53.2 | 25.3 | 9.1 | 12.3 | | Apr 2024 | 51.4 | 26.4 | 8.8 | 13.5 | | Mar 2024 | 51.4 | 26.3 | 8.9 | 13.4 | | Feb 2024 | 60.9 | 20.7 | 7.8 | 10.6 | | Jan 2024 | 48.8 | 27.1 | 9.3 | 14.7 | | Dec 2023 | 46.0 | 28.7 | 10.7 | 14.7 | | Nov 2023 | 46.2 | 27.4 | 11.3 | 15.1 | | Oct 2023 | 47.2 | 24.4 | 12.2 | 16.3 | | Sep 2023 | 43.4 | 27.3 | 12.1 | 17.2 | | Aug 2023 | 44.0 | 26.0 | 12.0 | 18.0 | | Jul 2023 | 44.6 | 21.5 | 12.4 | 21.5 | | Jun 2023 | 42.4 | 22.8 | 13.0 | 21.7 | | May 2023 | 38.8 | 23.5 | 14.1 | 23.5 | | Apr 2023 | 42.5 | 23.0 | 13.3 | 21.2 | | Mar 2023 | 40.4 | 22.5 | 13.5 | 23.6 | | Feb 2023 | 40.9 | 20.4 | 12.9 | 25.8 | | Jan 2023 | 39.5 | 17.5 | 13.2 | 29.8 | | Dec 2022 | 35.7 | 19.0 | 15.5 | 29.8 | | Nov 2022 | 33.3 | 20.5 | 12.8 | 33.3 | | Oct 2022 | 30.3 | 22.2 | 13.1 | 34.3 | | Sep 2022 | 28.3 | 28.3 | 12.0 | 31.5 | | Aug 2022 | 32.6 | 19.8 | 14.0 | 33.7 | | Jul 2022 | 32.7 | 19.1 | 13.6 | 34.5 | | Jun 2022 | 33.7 | 19.1 | 13.5 | 33.7 | | May 2022 | 24.1 | 26.7 | 17.2 | 31.9 | | Apr 2022 | 23.5 | 23.5 | 18.8 | 34.1 | | Mar 2022 | 24.0 | 24.0 | 16.0 | 36.0 | | Feb 2022 | 22.4 | 28.4 | 17.9 | 31.3 | | Jan 2022 | 23.8 | 27.4 | 17.9 | 31.0 | | Dec 2021 | 15.4 | 20.5 | 23.1 | 41.0 | | Nov 2021 | 0.0 | 15.4 | 23.1 | 61.5 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Quick commerce, October 2026, colleviate.com/blog/share-of-search-quick-commerce-india.html ## Questions **Who has the highest share of search in quick commerce in India?** Blinkit, with 56.4% of quick-commerce brand search over the three months to September 2026, ahead of Zepto at 28.0%. It has led every month since February 2025. **How often has the lead changed?** The lead has changed hands 5 times in 59 months. Counting every month since November 2021: Blinkit for 44 months, BigBasket for 11 months and Zepto for 4 months. **Is share of search the same as market share in quick commerce?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. Eternal, Blinkit's parent, and Swiggy, which owns Instamart, report quick-commerce gross order value in their quarterly results. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Quick commerce, October 2026, colleviate.com/blog/share-of-search-quick-commerce-india.html. --- # Swiggy vs Zomato: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-food-delivery-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Swiggy has the largest share of food-delivery company brand search in India: 50.9% over the three months to September 2026, against Zomato at 49.1%. It has led every month since July 2026. Over the past year Swiggy gained 1.6 points on Zomato. Two brands, one category, and five years in which neither has pulled clear in search. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/food-delivery.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads food delivery search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | Swiggy | 38.5% | 49.3% | 50.9% | +1.6 pts | | Zomato | 61.5% | 50.7% | 49.1% | -1.6 pts | SwiggyZomato Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed Zomato held 61.5% at the start. The two are now level, 50.9 to 49.1 in Swiggy's favour, and the lead has changed hands repeatedly since November 2024. Searches for Swiggy include people looking for Instamart, so this is interest in the company brands rather than in food delivery alone. ## Who has led, and for how long The lead has changed hands 9 times in 59 months. Counting every month since November 2021: Zomato for 45 months and Swiggy for 14 months. | Leader | From | To | Months | Public event that month | | Zomato | Nov 2021 | Oct 2024 | 36 | | | Swiggy | Nov 2024 | Nov 2024 | 1 | The month of its stock market listing | | Zomato | Dec 2024 | Apr 2025 | 5 | | | Swiggy | May 2025 | Jun 2025 | 2 | | | Zomato | Jul 2025 | Aug 2025 | 2 | | | Swiggy | Sep 2025 | Dec 2025 | 4 | | | Zomato | Jan 2026 | Jan 2026 | 1 | | | Swiggy | Feb 2026 | May 2026 | 4 | | | Zomato | Jun 2026 | Jun 2026 | 1 | | | Swiggy | Jul 2026 | Sep 2026 | 3 | | ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - November 2024: Zomato -10.9 points, to 44.9%. - November 2024: Swiggy +10.9 points, to 55.1%, the month of its stock market listing. - December 2024: Zomato +10.3 points, to 55.2%. Each brand's single highest week of search interest: Swiggy in November 2024 (the month of its stock market listing) and Zomato in January 2022. ## Search share and market share Eternal, Zomato's parent, and Swiggy both report food-delivery gross order value every quarter, which makes this the one category in the index where search share can be checked against audited figures for every brand in the set. Remember that searches for Swiggy include people looking for Instamart, so the search side is wider than food delivery alone. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “Swiggy” and “Zomato”. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | Swiggy | Zomato | | Sep 2026 | 51.5 | 48.5 | | Aug 2026 | 50.8 | 49.2 | | Jul 2026 | 50.4 | 49.6 | | Jun 2026 | 49.8 | 50.2 | | May 2026 | 50.8 | 49.2 | | Apr 2026 | 50.2 | 49.8 | | Mar 2026 | 50.6 | 49.4 | | Feb 2026 | 52.2 | 47.8 | | Jan 2026 | 47.3 | 52.7 | | Dec 2025 | 50.5 | 49.5 | | Nov 2025 | 50.9 | 49.1 | | Oct 2025 | 51.7 | 48.3 | | Sep 2025 | 50.6 | 49.4 | | Aug 2025 | 49.6 | 50.4 | | Jul 2025 | 47.8 | 52.2 | | Jun 2025 | 51.4 | 48.6 | | May 2025 | 51.5 | 48.5 | | Apr 2025 | 46.8 | 53.2 | | Mar 2025 | 43.9 | 56.1 | | Feb 2025 | 44.2 | 55.8 | | Jan 2025 | 40.5 | 59.5 | | Dec 2024 | 44.8 | 55.2 | | Nov 2024 | 55.1 | 44.9 | | Oct 2024 | 44.2 | 55.8 | | Sep 2024 | 42.3 | 57.7 | | Aug 2024 | 40.4 | 59.6 | | Jul 2024 | 40.9 | 59.1 | | Jun 2024 | 42.7 | 57.3 | | May 2024 | 41.6 | 58.4 | | Apr 2024 | 42.5 | 57.5 | | Mar 2024 | 40.5 | 59.5 | | Feb 2024 | 39.5 | 60.5 | | Jan 2024 | 40.5 | 59.5 | | Dec 2023 | 39.8 | 60.2 | | Nov 2023 | 37.7 | 62.3 | | Oct 2023 | 39.3 | 60.7 | | Sep 2023 | 40.6 | 59.4 | | Aug 2023 | 39.0 | 61.0 | | Jul 2023 | 41.2 | 58.8 | | Jun 2023 | 40.5 | 59.5 | | May 2023 | 41.0 | 59.0 | | Apr 2023 | 41.9 | 58.1 | | Mar 2023 | 42.8 | 57.2 | | Feb 2023 | 38.3 | 61.7 | | Jan 2023 | 38.8 | 61.2 | | Dec 2022 | 40.3 | 59.7 | | Nov 2022 | 37.2 | 62.8 | | Oct 2022 | 39.0 | 61.0 | | Sep 2022 | 38.3 | 61.7 | | Aug 2022 | 35.9 | 64.1 | | Jul 2022 | 33.1 | 66.9 | | Jun 2022 | 38.0 | 62.0 | | May 2022 | 36.6 | 63.4 | | Apr 2022 | 41.2 | 58.8 | | Mar 2022 | 37.4 | 62.6 | | Feb 2022 | 32.3 | 67.7 | | Jan 2022 | 34.5 | 65.5 | | Dec 2021 | 42.7 | 57.3 | | Nov 2021 | 38.4 | 61.6 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Food delivery (company brands), October 2026, colleviate.com/blog/share-of-search-food-delivery-india.html ## Questions **Who has the highest share of search in food delivery in India?** Swiggy, with 50.9% of food-delivery company brand search over the three months to September 2026, ahead of Zomato at 49.1%. It has led every month since July 2026. **How often has the lead changed?** The lead has changed hands 9 times in 59 months. Counting every month since November 2021: Zomato for 45 months and Swiggy for 14 months. **Is share of search the same as market share in food delivery?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. Eternal, Zomato's parent, and Swiggy both report food-delivery gross order value every quarter, which makes this the one category in the index where search share can be checked against audited figures for every brand in the set. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Food delivery (company brands), October 2026, colleviate.com/blog/share-of-search-food-delivery-india.html. --- # PhonePe vs Paytm vs Google Pay: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-upi-payments-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** PhonePe has the largest share of UPI payment app brand search in India: 37.0% over the three months to September 2026, against Paytm at 35.6%. It has led every month since May 2026. Over the past year Google Pay gained most, +3.7 points, while Paytm lost most, -3.6. UPI apps are opened every day and searched for far less, which makes this the category where search share and actual use part company most. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/upi-payments.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads UPI apps search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | PhonePe | 13.6% | 36.5% | 37.0% | +0.5 pts | | Google Pay | 15.8% | 17.8% | 21.5% | +3.7 pts | | Paytm | 67.8% | 39.2% | 35.6% | -3.6 pts | | CRED | 2.7% | 6.5% | 5.9% | -0.6 pts | PhonePeGoogle PayPaytmCRED Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed Paytm fell from 67.8% to 35.6%. PhonePe rose from 13.6% to 37.0%, first overtook Paytm in October 2025 and has led since May 2026. Paytm's own peak, 75.6% in February 2024, was the month the RBI restricted Paytm Payments Bank: attention, not demand. Google Pay holds 21.5% and CRED 5.9%. Search share is not payment share: Paytm's searches include its share price and its banking news, and an app used daily from the home screen rarely needs searching for. ## Who has led, and for how long The lead has changed hands 5 times in 59 months. Counting every month since November 2021: Paytm for 49 months and PhonePe for 10 months. | Leader | From | To | Months | Public event that month | | Paytm | Nov 2021 | Sep 2025 | 47 | | | PhonePe | Oct 2025 | Oct 2025 | 1 | | | Paytm | Nov 2025 | Nov 2025 | 1 | | | PhonePe | Dec 2025 | Mar 2026 | 4 | | | Paytm | Apr 2026 | Apr 2026 | 1 | | | PhonePe | May 2026 | Sep 2026 | 5 | | ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - December 2021: Paytm -13.2 points, to 62.4%. - March 2024: Paytm -13.0 points, to 62.6%. - April 2024: Paytm -8.8 points, to 53.8%. Each brand's single highest week of search interest: PhonePe in July 2024, Google Pay in February 2022, Paytm in November 2021 and CRED in June 2026. ## Search share and market share NPCI publishes the monthly volume and value of UPI transactions by app, the cleanest market-share series for any category in the index. It is also the clearest case of search share and use parting company: an app opened from the home screen every day rarely needs searching for, and Paytm's searches include its share price and its banking news. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “PhonePe”, “Google Pay”, “Paytm” and “CRED”. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | PhonePe | Google Pay | Paytm | CRED | | Sep 2026 | 37.2 | 22.6 | 34.3 | 5.8 | | Aug 2026 | 36.8 | 21.1 | 36.3 | 5.8 | | Jul 2026 | 36.9 | 20.8 | 36.2 | 6.2 | | Jun 2026 | 36.6 | 22.1 | 32.1 | 9.2 | | May 2026 | 37.1 | 20.5 | 35.8 | 6.6 | | Apr 2026 | 34.6 | 22.3 | 36.9 | 6.2 | | Mar 2026 | 37.7 | 21.6 | 34.7 | 6.0 | | Feb 2026 | 37.2 | 20.9 | 35.7 | 6.2 | | Jan 2026 | 38.1 | 20.9 | 35.3 | 5.8 | | Dec 2025 | 37.6 | 22.0 | 34.8 | 5.7 | | Nov 2025 | 36.3 | 20.8 | 36.9 | 6.0 | | Oct 2025 | 37.1 | 19.7 | 37.1 | 6.1 | | Sep 2025 | 36.4 | 17.8 | 38.0 | 7.8 | | Aug 2025 | 36.2 | 17.8 | 40.2 | 5.7 | | Jul 2025 | 37.0 | 17.8 | 39.3 | 5.9 | | Jun 2025 | 36.2 | 18.4 | 38.7 | 6.7 | | May 2025 | 37.0 | 17.3 | 39.4 | 6.3 | | Apr 2025 | 36.7 | 18.0 | 39.1 | 6.2 | | Mar 2025 | 36.2 | 17.5 | 40.1 | 6.2 | | Feb 2025 | 37.0 | 16.7 | 40.6 | 5.8 | | Jan 2025 | 34.4 | 16.2 | 43.8 | 5.6 | | Dec 2024 | 33.2 | 17.1 | 44.9 | 4.9 | | Nov 2024 | 31.6 | 16.8 | 45.8 | 5.8 | | Oct 2024 | 31.2 | 17.6 | 46.6 | 4.5 | | Sep 2024 | 32.0 | 16.2 | 45.9 | 5.9 | | Aug 2024 | 30.9 | 16.0 | 48.0 | 5.1 | | Jul 2024 | 31.1 | 15.0 | 48.9 | 5.0 | | Jun 2024 | 27.9 | 15.4 | 51.2 | 5.5 | | May 2024 | 24.6 | 15.6 | 55.1 | 4.8 | | Apr 2024 | 24.6 | 15.2 | 53.8 | 6.4 | | Mar 2024 | 20.4 | 12.9 | 62.6 | 4.1 | | Feb 2024 | 13.0 | 8.9 | 75.6 | 2.5 | | Jan 2024 | 16.4 | 11.2 | 69.5 | 3.0 | | Dec 2023 | 18.9 | 14.9 | 62.5 | 3.6 | | Nov 2023 | 22.4 | 16.7 | 56.8 | 4.2 | | Oct 2023 | 21.7 | 16.8 | 57.0 | 4.5 | | Sep 2023 | 21.7 | 15.9 | 58.2 | 4.2 | | Aug 2023 | 21.0 | 15.6 | 59.5 | 3.9 | | Jul 2023 | 21.4 | 16.1 | 58.5 | 4.0 | | Jun 2023 | 20.1 | 16.7 | 59.3 | 3.9 | | May 2023 | 18.4 | 16.3 | 61.2 | 4.1 | | Apr 2023 | 19.3 | 17.3 | 59.0 | 4.4 | | Mar 2023 | 18.1 | 17.2 | 60.8 | 3.9 | | Feb 2023 | 18.5 | 16.2 | 61.3 | 4.1 | | Jan 2023 | 18.1 | 15.9 | 61.7 | 4.3 | | Dec 2022 | 17.8 | 17.8 | 60.7 | 3.7 | | Nov 2022 | 17.3 | 17.3 | 61.1 | 4.4 | | Oct 2022 | 18.9 | 18.9 | 57.8 | 4.4 | | Sep 2022 | 17.3 | 17.3 | 61.8 | 3.6 | | Aug 2022 | 18.1 | 18.1 | 60.0 | 3.7 | | Jul 2022 | 17.7 | 18.5 | 60.1 | 3.7 | | Jun 2022 | 16.9 | 17.7 | 61.9 | 3.5 | | May 2022 | 16.0 | 17.3 | 62.9 | 3.8 | | Apr 2022 | 15.4 | 16.9 | 62.7 | 5.0 | | Mar 2022 | 11.9 | 14.1 | 70.8 | 3.2 | | Feb 2022 | 15.2 | 17.4 | 64.5 | 2.8 | | Jan 2022 | 14.9 | 16.6 | 65.4 | 3.1 | | Dec 2021 | 16.0 | 18.6 | 62.4 | 3.0 | | Nov 2021 | 10.0 | 12.3 | 75.6 | 2.1 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: UPI payment apps, October 2026, colleviate.com/blog/share-of-search-upi-payments-india.html ## Questions **Who has the highest share of search in UPI apps in India?** PhonePe, with 37.0% of UPI payment app brand search over the three months to September 2026, ahead of Paytm at 35.6%. It has led every month since May 2026. **How often has the lead changed?** The lead has changed hands 5 times in 59 months. Counting every month since November 2021: Paytm for 49 months and PhonePe for 10 months. **Is share of search the same as market share in UPI apps?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. NPCI publishes the monthly volume and value of UPI transactions by app, the cleanest market-share series for any category in the index. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: UPI payment apps, October 2026, colleviate.com/blog/share-of-search-upi-payments-india.html. --- # Groww vs Zerodha vs Angel One: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-stockbroking-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Groww has the largest share of stockbroking brand search in India: 40.1% over the three months to September 2026, against Zerodha at 30.2%. It has led every month since September 2025. Over the past year Groww gained most, +7.0 points, while Zerodha lost most, -3.3. A generation of first-time investors arrived after 2020, and branded search records which platform they went looking for. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/stockbroking.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads stockbroking search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | Zerodha | 61.2% | 33.5% | 30.2% | -3.3 pts | | Groww | 14.4% | 33.1% | 40.1% | +7.0 pts | | Upstox | 21.2% | 10.2% | 8.6% | -1.6 pts | | Angel One | 3.2% | 23.2% | 21.2% | -2.0 pts | ZerodhaGrowwUpstoxAngel One Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed Zerodha's share fell from 61.2% to 30.2%, while Groww rose from 14.4% to 40.1% and Angel One from 3.2% to 21.2%. Groww first led in January 2024 and has led every month since September 2025; its spike to 52.1% in November 2025 was the month of its stock market listing. Dhan is left out of the share: the word also means wealth in Hindi and appears inside other names, which would overstate it. ## Who has led, and for how long The lead has changed hands 7 times in 59 months. Counting every month since November 2021: Zerodha for 39 months and Groww for 20 months. | Leader | From | To | Months | Public event that month | | Zerodha | Nov 2021 | Dec 2023 | 26 | | | Groww | Jan 2024 | Jan 2024 | 1 | | | Zerodha | Feb 2024 | Oct 2024 | 9 | | | Groww | Nov 2024 | Feb 2025 | 4 | | | Zerodha | Mar 2025 | Apr 2025 | 2 | | | Groww | May 2025 | Jun 2025 | 2 | | | Zerodha | Jul 2025 | Aug 2025 | 2 | | | Groww | Sep 2025 | Sep 2026 | 13 | | ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - November 2025: Groww +16.6 points, to 52.1%, the month of its stock market listing. - December 2025: Groww -11.8 points, to 40.3%. - November 2025: Zerodha -9.4 points, to 24.8%. Each brand's single highest week of search interest: Zerodha in February 2024, Groww in November 2025 (the month of its stock market listing), Upstox in January 2022 and Angel One in August 2024. ## Search share and market share NSE publishes active client numbers by broker, and Angel One and Groww, both listed, report client numbers in their results. Active clients, rather than new accounts, are the closer match to what share of search tracks: people choosing a platform, or coming back to it. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “Zerodha”, “Groww”, “Upstox” and “Angel One”. Dhan is left out of the share: also the Hindi word for wealth, and part of other names such as Jan Dhan. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | Zerodha | Groww | Upstox | Angel One | | Sep 2026 | 30.7 | 40.8 | 8.8 | 19.7 | | Aug 2026 | 29.0 | 40.5 | 8.4 | 22.0 | | Jul 2026 | 30.8 | 39.0 | 8.5 | 21.8 | | Jun 2026 | 30.0 | 37.9 | 9.2 | 22.9 | | May 2026 | 28.4 | 39.7 | 8.7 | 23.2 | | Apr 2026 | 28.7 | 39.9 | 8.7 | 22.7 | | Mar 2026 | 29.5 | 39.3 | 8.8 | 22.4 | | Feb 2026 | 28.7 | 40.8 | 8.3 | 22.2 | | Jan 2026 | 30.4 | 40.2 | 8.1 | 21.2 | | Dec 2025 | 30.5 | 40.3 | 8.6 | 20.7 | | Nov 2025 | 24.8 | 52.1 | 7.0 | 16.1 | | Oct 2025 | 34.2 | 35.5 | 9.5 | 20.8 | | Sep 2025 | 33.5 | 34.1 | 9.7 | 22.7 | | Aug 2025 | 33.6 | 32.4 | 10.4 | 23.6 | | Jul 2025 | 33.5 | 32.7 | 10.5 | 23.3 | | Jun 2025 | 32.4 | 34.4 | 9.9 | 23.3 | | May 2025 | 32.1 | 34.3 | 10.2 | 23.4 | | Apr 2025 | 34.0 | 33.0 | 10.1 | 22.9 | | Mar 2025 | 33.6 | 32.9 | 10.6 | 22.9 | | Feb 2025 | 33.1 | 34.0 | 10.3 | 22.6 | | Jan 2025 | 32.3 | 34.3 | 10.0 | 23.4 | | Dec 2024 | 33.1 | 34.5 | 10.0 | 22.4 | | Nov 2024 | 32.9 | 34.7 | 10.4 | 22.0 | | Oct 2024 | 33.4 | 32.8 | 10.1 | 23.8 | | Sep 2024 | 35.6 | 33.0 | 9.6 | 21.7 | | Aug 2024 | 34.0 | 33.3 | 9.7 | 23.0 | | Jul 2024 | 37.1 | 31.3 | 10.3 | 21.3 | | Jun 2024 | 36.1 | 32.9 | 10.8 | 20.3 | | May 2024 | 35.7 | 31.8 | 10.3 | 22.2 | | Apr 2024 | 36.5 | 31.1 | 10.7 | 21.7 | | Mar 2024 | 35.7 | 31.6 | 11.1 | 21.7 | | Feb 2024 | 39.9 | 30.3 | 10.9 | 18.9 | | Jan 2024 | 35.3 | 35.8 | 10.6 | 18.2 | | Dec 2023 | 40.9 | 29.9 | 11.6 | 17.6 | | Nov 2023 | 41.9 | 28.0 | 11.9 | 18.1 | | Oct 2023 | 40.8 | 27.6 | 12.2 | 19.4 | | Sep 2023 | 43.5 | 28.2 | 12.5 | 15.7 | | Aug 2023 | 44.1 | 26.9 | 12.9 | 16.0 | | Jul 2023 | 47.0 | 25.2 | 12.9 | 14.9 | | Jun 2023 | 48.6 | 24.6 | 13.0 | 13.8 | | May 2023 | 48.3 | 24.2 | 13.8 | 13.8 | | Apr 2023 | 49.9 | 22.7 | 13.6 | 13.9 | | Mar 2023 | 49.0 | 22.1 | 12.8 | 16.1 | | Feb 2023 | 51.1 | 24.3 | 12.8 | 11.8 | | Jan 2023 | 53.2 | 23.6 | 12.7 | 10.4 | | Dec 2022 | 53.2 | 23.9 | 13.5 | 9.4 | | Nov 2022 | 55.6 | 21.9 | 13.6 | 8.9 | | Oct 2022 | 56.5 | 20.9 | 14.2 | 8.4 | | Sep 2022 | 59.2 | 19.7 | 14.4 | 6.7 | | Aug 2022 | 60.8 | 19.6 | 13.6 | 5.9 | | Jul 2022 | 62.1 | 17.9 | 14.3 | 5.7 | | Jun 2022 | 62.6 | 16.5 | 15.2 | 5.7 | | May 2022 | 60.4 | 17.4 | 16.3 | 5.8 | | Apr 2022 | 59.2 | 17.0 | 18.1 | 5.8 | | Mar 2022 | 59.5 | 16.8 | 18.7 | 5.1 | | Feb 2022 | 60.6 | 16.1 | 18.8 | 4.4 | | Jan 2022 | 60.5 | 16.6 | 19.2 | 3.7 | | Dec 2021 | 61.4 | 12.9 | 22.6 | 3.1 | | Nov 2021 | 61.7 | 13.7 | 21.8 | 2.8 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Stockbroking, October 2026, colleviate.com/blog/share-of-search-stockbroking-india.html ## Questions **Who has the highest share of search in stockbroking in India?** Groww, with 40.1% of stockbroking brand search over the three months to September 2026, ahead of Zerodha at 30.2%. It has led every month since September 2025. **How often has the lead changed?** The lead has changed hands 7 times in 59 months. Counting every month since November 2021: Zerodha for 39 months and Groww for 20 months. **Is share of search the same as market share in stockbroking?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. NSE publishes active client numbers by broker, and Angel One and Groww, both listed, report client numbers in their results. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Stockbroking, October 2026, colleviate.com/blog/share-of-search-stockbroking-india.html. --- # Ather vs Ola Electric vs TVS iQube: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-electric-two-wheelers-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Ather has the largest share of electric two-wheeler brand search in India: 62.2% over the three months to September 2026, against Ola Electric at 14.7%. It has led every month since September 2025. Over the past year Ather gained most, +24.1 points, while Ola Electric lost most, -24.4. The largest shift anywhere in the index: the early leader in electric scooter search lost most of its share to a rival. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/electric-two-wheelers.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads electric two-wheelers search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | Ola Electric | 48.7% | 39.1% | 14.7% | -24.4 pts | | Ather | 35.2% | 38.1% | 62.2% | +24.1 pts | | TVS iQube | 7.6% | 13.7% | 13.5% | -0.2 pts | | Bajaj Chetak | 8.5% | 9.1% | 9.5% | +0.4 pts | Ola ElectricAtherTVS iQubeBajaj Chetak Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed Ola Electric held 48.7% of the set's branded search at the start and 14.7% now. Ather rose from 35.2% to 62.2% and has led every month since September 2025. Ola Electric's peak search interest came in August 2024, the month of its stock market listing, a reminder that attention and buying intent are not the same thing. TVS iQube (13.5%) and Bajaj Chetak (9.5%) have held steadier. ## Who has led, and for how long The lead has changed hands 13 times in 59 months. Counting every month since November 2021: Ather for 37 months and Ola Electric for 22 months. The eight most recent spells in the lead: | Leader | From | To | Months | Public event that month | | Ola Electric | Dec 2023 | Dec 2023 | 1 | | | Ather | Jan 2024 | Jan 2024 | 1 | | | Ola Electric | Feb 2024 | Feb 2024 | 1 | | | Ather | Mar 2024 | Jun 2024 | 4 | | | Ola Electric | Jul 2024 | Mar 2025 | 9 | | | Ather | Apr 2025 | Jun 2025 | 3 | | | Ola Electric | Jul 2025 | Aug 2025 | 2 | | | Ather | Sep 2025 | Sep 2026 | 13 | | ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - August 2024: Ola Electric +27.7 points, to 75.7%, the month of its stock market listing. - April 2025: Ather +19.6 points, to 52.8%. - September 2024: Ola Electric -18.0 points, to 57.7%. Each brand's single highest week of search interest: Ola Electric in August 2024 (the month of its stock market listing), Ather in August 2026, TVS iQube in May 2022 and Bajaj Chetak in December 2024. ## Search share and market share The government's VAHAN dashboard records new vehicle registrations by maker, month by month, which is the closest public read of sales. When search share runs well ahead of registration share, it can mean interest that the product, the price or the dealer network is not yet converting. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “Ola Electric”, “Ather”, “TVS iQube” and “Bajaj Chetak”. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | Ola Electric | Ather | TVS iQube | Bajaj Chetak | | Sep 2026 | 15.1 | 62.7 | 13.1 | 9.1 | | Aug 2026 | 13.4 | 66.1 | 12.3 | 8.1 | | Jul 2026 | 15.7 | 57.8 | 15.2 | 11.3 | | Jun 2026 | 21.1 | 51.1 | 15.9 | 11.9 | | May 2026 | 21.9 | 50.0 | 16.8 | 11.3 | | Apr 2026 | 29.7 | 46.4 | 13.5 | 10.4 | | Mar 2026 | 21.7 | 50.8 | 15.9 | 11.6 | | Feb 2026 | 31.2 | 44.6 | 14.0 | 10.2 | | Jan 2026 | 29.2 | 44.1 | 14.3 | 12.4 | | Dec 2025 | 39.0 | 39.6 | 13.4 | 7.9 | | Nov 2025 | 33.2 | 45.0 | 13.9 | 7.9 | | Oct 2025 | 33.5 | 41.9 | 15.3 | 9.4 | | Sep 2025 | 33.8 | 41.2 | 15.3 | 9.7 | | Aug 2025 | 43.4 | 36.9 | 12.0 | 7.7 | | Jul 2025 | 40.0 | 36.1 | 13.9 | 10.0 | | Jun 2025 | 34.7 | 39.2 | 14.1 | 12.1 | | May 2025 | 31.9 | 47.0 | 10.8 | 10.2 | | Apr 2025 | 28.1 | 52.8 | 9.6 | 9.6 | | Mar 2025 | 44.7 | 33.2 | 10.1 | 12.1 | | Feb 2025 | 44.4 | 34.6 | 9.8 | 11.1 | | Jan 2025 | 46.2 | 33.1 | 8.9 | 11.8 | | Dec 2024 | 47.5 | 30.0 | 9.2 | 13.4 | | Nov 2024 | 53.8 | 29.3 | 9.1 | 7.7 | | Oct 2024 | 59.2 | 22.2 | 10.0 | 8.7 | | Sep 2024 | 57.7 | 25.7 | 8.8 | 7.8 | | Aug 2024 | 75.7 | 16.1 | 4.5 | 3.8 | | Jul 2024 | 48.0 | 32.4 | 11.0 | 8.7 | | Jun 2024 | 34.4 | 40.3 | 14.9 | 10.4 | | May 2024 | 37.7 | 38.6 | 16.7 | 7.0 | | Apr 2024 | 32.4 | 49.3 | 11.5 | 6.8 | | Mar 2024 | 34.8 | 39.1 | 17.4 | 8.7 | | Feb 2024 | 39.3 | 37.0 | 14.8 | 8.9 | | Jan 2024 | 31.6 | 39.1 | 18.0 | 11.3 | | Dec 2023 | 36.9 | 35.7 | 16.6 | 10.8 | | Nov 2023 | 33.9 | 36.3 | 19.4 | 10.5 | | Oct 2023 | 37.1 | 32.9 | 18.6 | 11.4 | | Sep 2023 | 39.3 | 35.2 | 18.0 | 7.4 | | Aug 2023 | 47.1 | 29.0 | 16.1 | 7.7 | | Jul 2023 | 36.6 | 37.3 | 19.0 | 7.2 | | Jun 2023 | 35.4 | 36.2 | 20.0 | 8.5 | | May 2023 | 33.1 | 38.7 | 20.4 | 7.7 | | Apr 2023 | 32.3 | 38.5 | 21.1 | 8.1 | | Mar 2023 | 33.6 | 35.6 | 22.6 | 8.2 | | Feb 2023 | 32.8 | 39.4 | 21.2 | 6.6 | | Jan 2023 | 30.7 | 46.4 | 17.7 | 5.2 | | Dec 2022 | 36.9 | 38.5 | 18.0 | 6.6 | | Nov 2022 | 35.3 | 39.5 | 18.5 | 6.7 | | Oct 2022 | 33.3 | 43.5 | 17.2 | 5.9 | | Sep 2022 | 32.1 | 40.2 | 20.5 | 7.1 | | Aug 2022 | 37.3 | 37.3 | 18.6 | 6.8 | | Jul 2022 | 30.2 | 43.4 | 18.6 | 7.8 | | Jun 2022 | 32.0 | 39.8 | 18.4 | 9.7 | | May 2022 | 30.3 | 42.8 | 20.0 | 6.9 | | Apr 2022 | 33.3 | 47.0 | 9.8 | 9.8 | | Mar 2022 | 46.7 | 36.4 | 7.3 | 9.7 | | Feb 2022 | 39.6 | 36.0 | 10.8 | 13.5 | | Jan 2022 | 44.8 | 37.9 | 7.8 | 9.5 | | Dec 2021 | 50.5 | 33.7 | 7.4 | 8.4 | | Nov 2021 | 50.8 | 33.9 | 7.6 | 7.6 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Electric two-wheelers, October 2026, colleviate.com/blog/share-of-search-electric-two-wheelers-india.html ## Questions **Who has the highest share of search in electric two-wheelers in India?** Ather, with 62.2% of electric two-wheeler brand search over the three months to September 2026, ahead of Ola Electric at 14.7%. It has led every month since September 2025. **How often has the lead changed?** The lead has changed hands 13 times in 59 months. Counting every month since November 2021: Ather for 37 months and Ola Electric for 22 months. **Is share of search the same as market share in electric two-wheelers?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. The government's VAHAN dashboard records new vehicle registrations by maker, month by month, which is the closest public read of sales. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Electric two-wheelers, October 2026, colleviate.com/blog/share-of-search-electric-two-wheelers-india.html. --- # Asian Paints vs Birla Opus vs Nerolac: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-paints-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Asian Paints has the largest share of decorative paint brand search in India: 72.1% over the three months to September 2026, against Birla Opus at 11.0%. It has led every month since the series began in November 2021. Over the past year Birla Opus gained most, +2.7 points, while Asian Paints lost most, -3.5. A category long dominated by one name, now with a new entrant in double digits. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/paints.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads paints search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. Indigo Paints is shown in the tables only, to keep the chart to four lines. | Brand | Start of series | A year ago | Last three months | Change in a year | | Asian Paints | 77.6% | 75.6% | 72.1% | -3.5 pts | | Berger Paints | 12.1% | 7.3% | 7.3% | +0.0 pts | | Nerolac | 6.4% | 7.1% | 7.6% | +0.5 pts | | Birla Opus | 0.0% | 8.3% | 11.0% | +2.7 pts | | Indigo Paints | 3.8% | 1.7% | 2.0% | +0.3 pts | Asian PaintsBerger PaintsNerolacBirla Opus Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed Asian Paints still holds 72.1%, down from 77.6%. Birla Opus first registered search interest in September 2023 and now holds 11.0%, ahead of Nerolac (7.6%) and Berger Paints (7.3%). Indigo Paints holds 2.0% and is shown in the table only, to keep the chart to four lines. ## Who has led, and for how long Asian Paints has led every month since the series began in November 2021, a lead that has never changed hands. ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - October 2023: Asian Paints +5.6 points, to 77.8%. - October 2023: Berger Paints -5.1 points, to 10.6%. - November 2024: Asian Paints +4.1 points, to 79.9%. Each brand's single highest week of search interest: Asian Paints in November 2024, Berger Paints in September 2023, Nerolac in October 2023, Birla Opus in September 2026 and Indigo Paints in October 2024. ## Search share and market share Asian Paints, Berger Paints, Kansai Nerolac and Indigo Paints are listed and report their revenue; Birla Opus sits inside Grasim Industries' results. Paint is bought rarely and often chosen with a contractor, so search share here reads the homeowner's consideration, not the whole decision. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “Asian Paints”, “Berger Paints”, “Nerolac”, “Birla Opus” and “Indigo Paints”. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | Asian Paints | Berger Paints | Nerolac | Birla Opus | Indigo Paints | | Sep 2026 | 72.3 | 6.8 | 7.7 | 11.5 | 1.7 | | Aug 2026 | 70.9 | 8.0 | 8.0 | 11.0 | 2.1 | | Jul 2026 | 73.2 | 7.1 | 7.1 | 10.4 | 2.2 | | Jun 2026 | 74.2 | 7.1 | 7.1 | 9.8 | 1.8 | | May 2026 | 74.1 | 6.9 | 7.2 | 10.0 | 1.7 | | Apr 2026 | 76.2 | 6.0 | 6.5 | 9.7 | 1.6 | | Mar 2026 | 76.9 | 6.2 | 6.2 | 8.6 | 2.1 | | Feb 2026 | 75.8 | 7.1 | 6.7 | 8.8 | 1.7 | | Jan 2026 | 76.2 | 7.1 | 6.7 | 8.3 | 1.7 | | Dec 2025 | 76.6 | 6.6 | 6.1 | 8.6 | 2.0 | | Nov 2025 | 75.7 | 6.8 | 6.5 | 8.3 | 2.7 | | Oct 2025 | 75.5 | 6.6 | 7.7 | 8.8 | 1.5 | | Sep 2025 | 75.6 | 6.3 | 7.7 | 8.9 | 1.5 | | Aug 2025 | 74.7 | 7.9 | 7.2 | 8.3 | 1.8 | | Jul 2025 | 76.4 | 7.7 | 6.4 | 7.7 | 1.8 | | Jun 2025 | 74.3 | 9.1 | 6.5 | 7.5 | 2.6 | | May 2025 | 76.6 | 7.7 | 6.1 | 7.3 | 2.3 | | Apr 2025 | 76.4 | 8.5 | 6.2 | 7.4 | 1.6 | | Mar 2025 | 76.8 | 7.7 | 6.2 | 7.1 | 2.1 | | Feb 2025 | 78.2 | 7.7 | 6.3 | 6.0 | 1.8 | | Jan 2025 | 78.6 | 7.6 | 5.9 | 5.5 | 2.4 | | Dec 2024 | 78.4 | 7.9 | 5.8 | 5.5 | 2.4 | | Nov 2024 | 79.9 | 7.0 | 5.5 | 4.9 | 2.6 | | Oct 2024 | 75.8 | 7.9 | 8.5 | 4.1 | 3.8 | | Sep 2024 | 75.6 | 9.6 | 8.0 | 3.6 | 3.1 | | Aug 2024 | 75.7 | 10.4 | 7.7 | 3.9 | 2.3 | | Jul 2024 | 75.2 | 9.4 | 8.7 | 4.3 | 2.4 | | Jun 2024 | 74.7 | 10.8 | 8.1 | 4.4 | 2.0 | | May 2024 | 77.8 | 10.9 | 7.3 | 1.5 | 2.5 | | Apr 2024 | 79.0 | 9.2 | 7.3 | 1.5 | 3.1 | | Mar 2024 | 80.1 | 8.1 | 7.5 | 1.4 | 2.9 | | Feb 2024 | 79.3 | 9.9 | 6.6 | 1.6 | 2.6 | | Jan 2024 | 81.1 | 8.5 | 7.0 | 0.0 | 3.3 | | Dec 2023 | 77.8 | 10.9 | 8.0 | 0.0 | 3.2 | | Nov 2023 | 77.5 | 10.9 | 7.9 | 0.0 | 3.7 | | Oct 2023 | 77.8 | 10.6 | 8.7 | 0.0 | 3.0 | | Sep 2023 | 72.2 | 15.7 | 8.6 | 0.3 | 3.2 | | Aug 2023 | 75.5 | 13.3 | 6.8 | 0.0 | 4.3 | | Jul 2023 | 76.4 | 10.2 | 9.3 | 0.0 | 4.2 | | Jun 2023 | 79.2 | 9.4 | 8.2 | 0.0 | 3.1 | | May 2023 | 76.6 | 9.7 | 9.4 | 0.0 | 4.3 | | Apr 2023 | 79.0 | 9.6 | 8.0 | 0.0 | 3.5 | | Mar 2023 | 79.1 | 10.3 | 7.1 | 0.0 | 3.6 | | Feb 2023 | 78.7 | 10.1 | 6.5 | 0.0 | 4.7 | | Jan 2023 | 80.3 | 10.7 | 6.1 | 0.0 | 2.9 | | Dec 2022 | 77.3 | 11.6 | 7.6 | 0.0 | 3.6 | | Nov 2022 | 76.5 | 11.9 | 7.1 | 0.0 | 4.5 | | Oct 2022 | 77.6 | 11.1 | 8.7 | 0.0 | 2.6 | | Sep 2022 | 76.5 | 11.3 | 9.3 | 0.0 | 3.0 | | Aug 2022 | 77.9 | 11.2 | 7.8 | 0.0 | 3.1 | | Jul 2022 | 76.6 | 12.7 | 7.7 | 0.0 | 3.0 | | Jun 2022 | 80.0 | 11.3 | 6.5 | 0.0 | 2.2 | | May 2022 | 78.8 | 11.5 | 6.7 | 0.0 | 2.9 | | Apr 2022 | 79.7 | 10.0 | 7.6 | 0.0 | 2.8 | | Mar 2022 | 79.0 | 11.2 | 6.8 | 0.0 | 3.0 | | Feb 2022 | 77.8 | 11.7 | 6.7 | 0.0 | 3.8 | | Jan 2022 | 80.4 | 10.3 | 6.1 | 0.0 | 3.1 | | Dec 2021 | 77.3 | 11.7 | 6.1 | 0.0 | 4.9 | | Nov 2021 | 75.2 | 14.3 | 7.1 | 0.0 | 3.4 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Decorative paints, October 2026, colleviate.com/blog/share-of-search-paints-india.html ## Questions **Who has the highest share of search in paints in India?** Asian Paints, with 72.1% of decorative paint brand search over the three months to September 2026, ahead of Birla Opus at 11.0%. It has led every month since the series began in November 2021. **How often has the lead changed?** Never. Asian Paints has led every month since the series began in November 2021. **Is share of search the same as market share in paints?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. Asian Paints, Berger Paints, Kansai Nerolac and Indigo Paints are listed and report their revenue; Birla Opus sits inside Grasim Industries' results. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Decorative paints, October 2026, colleviate.com/blog/share-of-search-paints-india.html. --- # IndiGo vs Air India: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-airlines-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** IndiGo has the largest share of domestic airline brand search in India: 55.6% over the three months to September 2026, against Air India at 37.8%. It has led every month since July 2025. Shares barely moved over the past year: no brand gained a full point. Airline search is driven by fares and news as much as by preference, which makes this the category that shows best why every spike needs a cause. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/airlines.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads airlines search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | IndiGo | 54.0% | 54.9% | 55.6% | +0.7 pts | | Air India | 33.5% | 37.6% | 37.8% | +0.2 pts | | Akasa Air | 0.2% | 3.1% | 3.0% | -0.1 pts | | SpiceJet | 12.2% | 4.3% | 3.6% | -0.7 pts | IndiGoAir IndiaAkasa AirSpiceJet Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed IndiGo leads with 55.6%, close to where it started. Air India rose from 33.5% to 37.8%. Its only month in the lead was June 2025, the month of the Ahmedabad crash, which is why every spike in share of search needs a cause attached before anyone reads it as demand. SpiceJet fell from 12.2% to 3.6%; Akasa Air grew from almost nothing to 3.0%. ## Who has led, and for how long The lead has changed hands 2 times in 59 months. Counting every month since November 2021: IndiGo for 58 months and Air India for 1 month. | Leader | From | To | Months | Public event that month | | IndiGo | Nov 2021 | May 2025 | 43 | | | Air India | Jun 2025 | Jun 2025 | 1 | The month of the Ahmedabad crash | | IndiGo | Jul 2025 | Sep 2026 | 15 | | ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - June 2025: Air India +17.9 points, to 57.6%, the month of the Ahmedabad crash. - June 2025: IndiGo -15.7 points, to 36.5%. - July 2025: IndiGo +14.7 points, to 51.2%. Each brand's single highest week of search interest: IndiGo in November 2025, Air India in June 2025 (the month of the Ahmedabad crash), Akasa Air in August 2022 and SpiceJet in November 2021. ## Search share and market share DGCA publishes domestic passenger numbers and market share by airline every month. Search share in this category is unusually sensitive to news: delays, incidents and fare sales all drive searches that have little to do with choosing an airline. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “IndiGo”, “Air India”, “Akasa Air” and “SpiceJet”. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | IndiGo | Air India | Akasa Air | SpiceJet | | Sep 2026 | 56.6 | 37.2 | 3.1 | 3.1 | | Aug 2026 | 54.0 | 38.6 | 2.8 | 4.5 | | Jul 2026 | 56.2 | 37.5 | 3.1 | 3.1 | | Jun 2026 | 55.0 | 39.3 | 2.9 | 2.9 | | May 2026 | 53.8 | 39.2 | 2.7 | 4.3 | | Apr 2026 | 52.6 | 38.3 | 3.2 | 5.8 | | Mar 2026 | 51.5 | 40.3 | 3.0 | 5.2 | | Feb 2026 | 54.3 | 38.3 | 2.3 | 5.1 | | Jan 2026 | 54.7 | 37.4 | 3.7 | 4.2 | | Dec 2025 | 64.3 | 29.3 | 2.7 | 3.7 | | Nov 2025 | 60.7 | 32.3 | 3.0 | 3.9 | | Oct 2025 | 56.4 | 34.3 | 4.7 | 4.7 | | Sep 2025 | 57.6 | 35.1 | 2.6 | 4.6 | | Aug 2025 | 56.0 | 34.8 | 4.3 | 4.8 | | Jul 2025 | 51.2 | 42.9 | 2.4 | 3.5 | | Jun 2025 | 36.5 | 57.6 | 2.9 | 2.9 | | May 2025 | 52.2 | 39.7 | 3.8 | 4.3 | | Apr 2025 | 55.4 | 36.3 | 3.6 | 4.8 | | Mar 2025 | 54.5 | 36.4 | 3.8 | 5.3 | | Feb 2025 | 55.6 | 35.4 | 4.5 | 4.5 | | Jan 2025 | 57.8 | 33.3 | 4.2 | 4.7 | | Dec 2024 | 56.1 | 34.3 | 4.3 | 5.2 | | Nov 2024 | 54.6 | 36.2 | 4.1 | 5.1 | | Oct 2024 | 57.0 | 33.5 | 4.5 | 5.0 | | Sep 2024 | 58.4 | 30.6 | 4.6 | 6.4 | | Aug 2024 | 59.0 | 30.6 | 4.6 | 5.8 | | Jul 2024 | 56.7 | 31.0 | 4.7 | 7.6 | | Jun 2024 | 55.3 | 34.0 | 4.8 | 5.9 | | May 2024 | 51.9 | 38.3 | 4.9 | 4.9 | | Apr 2024 | 55.5 | 35.5 | 3.2 | 5.8 | | Mar 2024 | 55.9 | 32.3 | 4.3 | 7.5 | | Feb 2024 | 55.0 | 32.5 | 5.0 | 7.5 | | Jan 2024 | 54.9 | 31.0 | 4.3 | 9.8 | | Dec 2023 | 55.5 | 30.5 | 3.6 | 10.5 | | Nov 2023 | 56.0 | 32.1 | 3.1 | 8.8 | | Oct 2023 | 56.1 | 32.9 | 3.5 | 7.5 | | Sep 2023 | 58.5 | 30.3 | 4.2 | 7.0 | | Aug 2023 | 58.7 | 30.7 | 4.0 | 6.7 | | Jul 2023 | 60.2 | 28.4 | 5.7 | 5.7 | | Jun 2023 | 57.9 | 30.3 | 5.3 | 6.6 | | May 2023 | 55.8 | 29.7 | 5.8 | 8.7 | | Apr 2023 | 56.9 | 31.2 | 4.4 | 7.5 | | Mar 2023 | 55.6 | 33.1 | 3.0 | 8.3 | | Feb 2023 | 55.6 | 32.7 | 2.6 | 9.2 | | Jan 2023 | 53.8 | 32.8 | 2.7 | 10.8 | | Dec 2022 | 56.1 | 30.4 | 2.7 | 10.8 | | Nov 2022 | 57.1 | 30.0 | 2.9 | 10.0 | | Oct 2022 | 57.6 | 29.1 | 3.5 | 9.9 | | Sep 2022 | 56.6 | 31.6 | 2.9 | 8.8 | | Aug 2022 | 54.2 | 29.9 | 6.2 | 9.7 | | Jul 2022 | 56.7 | 28.7 | 3.9 | 10.7 | | Jun 2022 | 55.5 | 31.4 | 0.7 | 12.4 | | May 2022 | 56.1 | 30.0 | 0.6 | 13.3 | | Apr 2022 | 55.1 | 33.3 | 0.0 | 11.6 | | Mar 2022 | 53.0 | 34.4 | 0.0 | 12.6 | | Feb 2022 | 53.0 | 35.8 | 0.0 | 11.3 | | Jan 2022 | 52.1 | 37.6 | 0.0 | 10.3 | | Dec 2021 | 55.6 | 31.7 | 0.7 | 12.0 | | Nov 2021 | 54.4 | 31.3 | 0.0 | 14.3 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Domestic airlines, October 2026, colleviate.com/blog/share-of-search-airlines-india.html ## Questions **Who has the highest share of search in airlines in India?** IndiGo, with 55.6% of domestic airline brand search over the three months to September 2026, ahead of Air India at 37.8%. It has led every month since July 2025. **How often has the lead changed?** The lead has changed hands 2 times in 59 months. Counting every month since November 2021: IndiGo for 58 months and Air India for 1 month. **Is share of search the same as market share in airlines?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. DGCA publishes domestic passenger numbers and market share by airline every month. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Domestic airlines, October 2026, colleviate.com/blog/share-of-search-airlines-india.html. --- # Jio vs Airtel vs Vi: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-telecom-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Jio has the largest share of prepaid recharge searches by network in India: 44.4% over the three months to September 2026, against Airtel at 34.9%. It has led every month since August 2024. Over the past year Airtel gained most, +2.3 points, while BSNL lost most, -4.3. Measured on recharge searches rather than brand names, this category reads what existing customers do as much as whom people choose. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/telecom.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads mobile networks search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | Jio | 37.6% | 43.2% | 44.4% | +1.2 pts | | Airtel | 37.3% | 32.6% | 34.9% | +2.3 pts | | Vi | 8.1% | 9.7% | 10.6% | +0.9 pts | | BSNL | 17.0% | 14.5% | 10.2% | -4.3 pts | JioAirtelViBSNL Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed Bare searches for Jio or Airtel also match streaming, banking and finance brands, so this category uses prepaid recharge searches instead. Jio leads with 44.4%, Airtel follows with 34.9%. BSNL briefly led in July 2024, the month the private operators raised tariffs, and has since fallen to 10.2%. Vi holds 10.6%. ## Who has led, and for how long The lead has changed hands 7 times in 59 months. Counting every month since November 2021: Jio for 46 months, Airtel for 12 months and BSNL for 1 month. | Leader | From | To | Months | Public event that month | | Airtel | Nov 2021 | Nov 2021 | 1 | | | Jio | Dec 2021 | May 2022 | 6 | | | Airtel | Jun 2022 | Jul 2022 | 2 | | | Jio | Aug 2022 | Sep 2022 | 2 | | | Airtel | Oct 2022 | Jun 2023 | 9 | | | Jio | Jul 2023 | Jun 2024 | 12 | | | BSNL | Jul 2024 | Jul 2024 | 1 | The month the private operators raised tariffs | | Jio | Aug 2024 | Sep 2026 | 26 | | ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - July 2024: BSNL +20.2 points, to 36.6%, the month the private operators raised tariffs. - August 2024: BSNL -13.5 points, to 23.1%. - July 2024: Jio -11.4 points, to 30.6%. Each brand's single highest week of search interest: Jio in June 2024, Airtel in November 2021, Vi in June 2024 and BSNL in July 2024 (the month the private operators raised tariffs). ## Search share and market share TRAI publishes subscriber numbers by operator every month. Because this category is measured on prepaid recharge searches, it reads existing customers' activity as much as switching, and a tariff change moves it more than a campaign does. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “Jio recharge”, “Airtel recharge”, “Vi recharge” and “BSNL recharge”. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | Jio | Airtel | Vi | BSNL | | Sep 2026 | 44.6 | 33.9 | 10.7 | 10.7 | | Aug 2026 | 43.2 | 36.2 | 10.2 | 10.5 | | Jul 2026 | 45.3 | 34.5 | 10.8 | 9.4 | | Jun 2026 | 45.4 | 34.4 | 10.6 | 9.6 | | May 2026 | 45.0 | 34.2 | 11.2 | 9.6 | | Apr 2026 | 44.0 | 34.4 | 11.5 | 10.0 | | Mar 2026 | 43.4 | 34.7 | 10.9 | 10.9 | | Feb 2026 | 42.9 | 34.2 | 10.7 | 12.2 | | Jan 2026 | 45.1 | 29.9 | 11.8 | 13.2 | | Dec 2025 | 46.1 | 29.5 | 10.6 | 13.8 | | Nov 2025 | 45.2 | 30.1 | 10.8 | 14.0 | | Oct 2025 | 43.1 | 32.0 | 10.7 | 14.2 | | Sep 2025 | 43.3 | 33.0 | 9.4 | 14.2 | | Aug 2025 | 43.0 | 32.1 | 10.0 | 15.0 | | Jul 2025 | 43.3 | 32.6 | 9.8 | 14.3 | | Jun 2025 | 43.5 | 34.0 | 10.2 | 12.2 | | May 2025 | 43.9 | 34.1 | 9.4 | 12.6 | | Apr 2025 | 43.0 | 33.2 | 10.3 | 13.6 | | Mar 2025 | 43.5 | 32.1 | 10.0 | 14.4 | | Feb 2025 | 42.7 | 32.9 | 9.8 | 14.5 | | Jan 2025 | 42.2 | 32.0 | 10.2 | 15.6 | | Dec 2024 | 42.1 | 31.2 | 9.9 | 16.8 | | Nov 2024 | 40.1 | 32.1 | 9.5 | 18.3 | | Oct 2024 | 40.3 | 33.5 | 8.6 | 17.6 | | Sep 2024 | 40.9 | 31.7 | 9.2 | 18.2 | | Aug 2024 | 37.5 | 30.3 | 9.1 | 23.1 | | Jul 2024 | 30.6 | 26.2 | 6.6 | 36.6 | | Jun 2024 | 42.0 | 32.1 | 9.5 | 16.4 | | May 2024 | 38.8 | 37.2 | 10.6 | 13.3 | | Apr 2024 | 38.5 | 37.5 | 10.4 | 13.5 | | Mar 2024 | 40.5 | 35.1 | 10.8 | 13.5 | | Feb 2024 | 39.8 | 34.0 | 11.0 | 15.2 | | Jan 2024 | 39.8 | 35.2 | 10.6 | 14.4 | | Dec 2023 | 40.1 | 35.5 | 11.0 | 13.5 | | Nov 2023 | 38.6 | 36.3 | 11.2 | 13.9 | | Oct 2023 | 38.8 | 35.9 | 11.2 | 14.1 | | Sep 2023 | 38.1 | 35.4 | 11.7 | 14.8 | | Aug 2023 | 37.2 | 34.3 | 13.6 | 14.9 | | Jul 2023 | 37.5 | 36.9 | 11.6 | 14.0 | | Jun 2023 | 36.4 | 36.9 | 11.9 | 14.8 | | May 2023 | 36.6 | 37.5 | 11.2 | 14.7 | | Apr 2023 | 36.0 | 37.6 | 10.9 | 15.5 | | Mar 2023 | 34.3 | 39.2 | 10.9 | 15.5 | | Feb 2023 | 33.2 | 41.2 | 10.6 | 15.0 | | Jan 2023 | 35.2 | 37.5 | 10.2 | 17.2 | | Dec 2022 | 36.0 | 36.3 | 10.9 | 16.9 | | Nov 2022 | 35.9 | 37.4 | 11.1 | 15.6 | | Oct 2022 | 36.9 | 37.2 | 9.9 | 16.1 | | Sep 2022 | 37.3 | 35.9 | 10.6 | 16.2 | | Aug 2022 | 36.4 | 35.4 | 10.1 | 18.2 | | Jul 2022 | 35.3 | 35.5 | 9.4 | 19.8 | | Jun 2022 | 36.1 | 36.4 | 10.1 | 17.4 | | May 2022 | 37.0 | 35.8 | 9.1 | 18.0 | | Apr 2022 | 36.9 | 35.4 | 8.9 | 18.8 | | Mar 2022 | 37.6 | 36.0 | 8.4 | 18.0 | | Feb 2022 | 38.2 | 35.1 | 8.8 | 17.9 | | Jan 2022 | 37.4 | 36.0 | 8.8 | 17.8 | | Dec 2021 | 37.4 | 36.2 | 8.0 | 18.4 | | Nov 2021 | 37.9 | 39.6 | 7.6 | 14.9 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Mobile networks (prepaid recharge), October 2026, colleviate.com/blog/share-of-search-telecom-india.html ## Questions **Who has the highest share of search in mobile networks in India?** Jio, with 44.4% of prepaid recharge searches by network over the three months to September 2026, ahead of Airtel at 34.9%. It has led every month since August 2024. **How often has the lead changed?** The lead has changed hands 7 times in 59 months. Counting every month since November 2021: Jio for 46 months, Airtel for 12 months and BSNL for 1 month. **Is share of search the same as market share in mobile networks?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. TRAI publishes subscriber numbers by operator every month. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Mobile networks (prepaid recharge), October 2026, colleviate.com/blog/share-of-search-telecom-india.html. --- # Nykaa vs Purplle: share of search in India, 2021 to 2026 URL: https://colleviate.com/blog/share-of-search-beauty-ecommerce-india.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Nykaa has the largest share of beauty e-commerce brand search in India: 91.4% over the three months to September 2026, against Purplle at 8.6%. It has led every month since the series began in November 2021. Shares barely moved over the past year: no brand gained a full point. One brand holds most of the branded search in online beauty, so the line worth watching is the challenger's. From the Colleviate Share of Search Index, with the data free to download. Download this category (CSV) (https://colleviate.com/data/sos/beauty-ecommerce.csv) Monthly, November 2021 to September 2026. Free to use with attribution. All nine categories (https://colleviate.com/data/colleviate-share-of-search-india.csv) ## Who leads beauty e-commerce search now Shares are the average of the last three months, set against the same three months a year earlier and the first three months of the series. | Brand | Start of series | A year ago | Last three months | Change in a year | | Nykaa | 95.5% | 91.1% | 91.4% | +0.3 pts | | Purplle | 4.5% | 8.9% | 8.6% | -0.3 pts | NykaaPurplle Share of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis. ## What changed Nykaa holds 91.4% of branded search against Purplle's 8.6%. Tira Beauty, Reliance's beauty platform, is left out because its volume sits below the level at which Google Trends reports a value. ## Who has led, and for how long Nykaa has led every month since the series began in November 2021, a lead that has never changed hands. ## The months that moved it The largest month-on-month changes in any brand's share. A spike needs a cause before it is read as demand: where a public, dated event explains one, it is named; the rest are reported as they are. - January 2025: Purplle -5.9 points, to 2.4%. - January 2025: Nykaa +5.9 points, to 97.6%. - February 2025: Purplle +5.6 points, to 8.0%. Each brand's single highest week of search interest: Nykaa in November 2021 and Purplle in October 2022. ## Search share and market share Nykaa's parent, FSN E-Commerce Ventures, is listed and reports beauty gross merchandise value every quarter. On a two-brand set this lopsided, the useful reading is the direction of the challenger's line, not the size of the leader's lead. This is attention, not sales: see share of search and market share (https://colleviate.com/blog/share-of-search-vs-market-share.html), and what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html) for how to read a number like this one. ## How this category is measured The search terms are “Nykaa” and “Purplle”. Tira Beauty is left out of the share: below the threshold at which Google Trends reports a value. They were pulled together, so they share one scale. Data retrieved 14 October 2026; the full method is on the index page (https://colleviate.com/blog/share-of-search-index-india.html#method). ## The monthly data Share of search by month, November 2021 to September 2026 (%) | Month | Nykaa | Purplle | | Sep 2026 | 90.5 | 9.5 | | Aug 2026 | 91.1 | 8.9 | | Jul 2026 | 92.5 | 7.5 | | Jun 2026 | 89.7 | 10.3 | | May 2026 | 91.2 | 8.8 | | Apr 2026 | 90.5 | 9.5 | | Mar 2026 | 95.7 | 4.3 | | Feb 2026 | 92.3 | 7.7 | | Jan 2026 | 90.7 | 9.3 | | Dec 2025 | 93.0 | 7.0 | | Nov 2025 | 93.8 | 6.2 | | Oct 2025 | 91.1 | 8.9 | | Sep 2025 | 90.0 | 10.0 | | Aug 2025 | 90.6 | 9.4 | | Jul 2025 | 92.6 | 7.4 | | Jun 2025 | 89.6 | 10.4 | | May 2025 | 91.3 | 8.7 | | Apr 2025 | 90.5 | 9.5 | | Mar 2025 | 90.6 | 9.4 | | Feb 2025 | 92.0 | 8.0 | | Jan 2025 | 97.6 | 2.4 | | Dec 2024 | 91.7 | 8.3 | | Nov 2024 | 96.6 | 3.4 | | Oct 2024 | 94.0 | 6.0 | | Sep 2024 | 92.8 | 7.2 | | Aug 2024 | 93.1 | 6.9 | | Jul 2024 | 93.7 | 6.3 | | Jun 2024 | 91.4 | 8.6 | | May 2024 | 92.3 | 7.7 | | Apr 2024 | 91.7 | 8.3 | | Mar 2024 | 86.2 | 13.8 | | Feb 2024 | 86.0 | 14.0 | | Jan 2024 | 91.5 | 8.5 | | Dec 2023 | 92.1 | 7.9 | | Nov 2023 | 93.8 | 6.2 | | Oct 2023 | 90.8 | 9.2 | | Sep 2023 | 91.7 | 8.3 | | Aug 2023 | 90.0 | 10.0 | | Jul 2023 | 92.9 | 7.1 | | Jun 2023 | 92.2 | 7.8 | | May 2023 | 92.5 | 7.5 | | Apr 2023 | 92.9 | 7.1 | | Mar 2023 | 92.3 | 7.7 | | Feb 2023 | 93.8 | 6.2 | | Jan 2023 | 93.8 | 6.2 | | Dec 2022 | 93.4 | 6.6 | | Nov 2022 | 95.9 | 4.1 | | Oct 2022 | 90.9 | 9.1 | | Sep 2022 | 93.0 | 7.0 | | Aug 2022 | 92.7 | 7.3 | | Jul 2022 | 93.3 | 6.7 | | Jun 2022 | 92.5 | 7.5 | | May 2022 | 93.8 | 6.2 | | Apr 2022 | 93.4 | 6.6 | | Mar 2022 | 94.1 | 5.9 | | Feb 2022 | 94.9 | 5.1 | | Jan 2022 | 94.3 | 5.7 | | Dec 2021 | 94.2 | 5.8 | | Nov 2021 | 97.9 | 2.1 | ## Cite this data Free to use, quote and republish under a Creative Commons Attribution 4.0 licence, with a link to this page. Colleviate Share of Search Index, India: Beauty e-commerce, October 2026, colleviate.com/blog/share-of-search-beauty-ecommerce-india.html ## Questions **Who has the highest share of search in beauty e-commerce in India?** Nykaa, with 91.4% of beauty e-commerce brand search over the three months to September 2026, ahead of Purplle at 8.6%. It has led every month since the series began in November 2021. **How often has the lead changed?** Never. Nykaa has led every month since the series began in November 2021. **Is share of search the same as market share in beauty e-commerce?** No. It measures attention: the share of branded searches, which tends to move before market share but is not sales. Nykaa's parent, FSN E-Commerce Ventures, is listed and reports beauty gross merchandise value every quarter. **Can I use this data?** Yes. It is published under a Creative Commons Attribution 4.0 licence. Download the CSV from this page and cite it as: Colleviate Share of Search Index, India: Beauty e-commerce, October 2026, colleviate.com/blog/share-of-search-beauty-ecommerce-india.html. --- # What is a good share of search? Benchmarks from nine Indian markets URL: https://colleviate.com/blog/what-is-a-good-share-of-search.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** There is no universal good share of search. In the nine Indian categories we track, the leading brand's share of branded search ranges from 37.0% (PhonePe, UPI apps) to 91.4% (Nykaa, beauty e-commerce). A good share of search is one above your share of the market and rising, measured against the brands your buyers actually compare you with. Everyone who starts tracking share of search asks the same question within a week: is our number good? The honest answer is that it depends on three things, and all three can be measured. ## Why there is no single good number Share of search is a share of a set you define. The same brand can hold 30% against five competitors and 60% against two, without a single search changing. So a number on its own says very little. It starts to mean something when it is read against three things: how many brands share the set, what your share of the market is, and which way it is moving. The Colleviate Share of Search Index (https://colleviate.com/blog/share-of-search-index-india.html) makes the first of those concrete, because it measures nine Indian categories the same way. These are the benchmarks it gives, as averages of the three months to September 2026. | Category | Brands | Leader, share | Lead over second | Even split | Concentration (HHI) | Changes of lead since Nov 2021 | | Beauty e-commerce (https://colleviate.com/blog/share-of-search-beauty-ecommerce-india.html) | 2 | Nykaa, 91.4% | 82.8 pts | 50% | 8,428 | 0 | | Paints (https://colleviate.com/blog/share-of-search-paints-india.html) | 5 | Asian Paints, 72.1% | 61.1 pts | 20% | 5,434 | 0 | | Electric two-wheelers (https://colleviate.com/blog/share-of-search-electric-two-wheelers-india.html) | 4 | Ather, 62.2% | 47.5 pts | 25% | 4,357 | 13 | | Quick commerce (https://colleviate.com/blog/share-of-search-quick-commerce-india.html) | 4 | Blinkit, 56.4% | 28.4 pts | 25% | 4,103 | 5 | | Airlines (https://colleviate.com/blog/share-of-search-airlines-india.html) | 4 | IndiGo, 55.6% | 17.8 pts | 25% | 4,542 | 2 | | Food delivery (https://colleviate.com/blog/share-of-search-food-delivery-india.html) | 2 | Swiggy, 50.9% | 1.8 pts | 50% | 5,002 | 9 | | Mobile networks (https://colleviate.com/blog/share-of-search-telecom-india.html) | 4 | Jio, 44.4% | 9.5 pts | 25% | 3,406 | 7 | | Stockbroking (https://colleviate.com/blog/share-of-search-stockbroking-india.html) | 4 | Groww, 40.1% | 9.9 pts | 25% | 3,043 | 7 | | UPI apps (https://colleviate.com/blog/share-of-search-upi-payments-india.html) | 4 | PhonePe, 37.0% | 1.4 pts | 25% | 3,133 | 5 | The even split is what each brand would hold if search were shared equally across the set. Concentration is the Herfindahl-Hirschman index, the sum of the squared shares: 10,000 means one brand has all the search, and an even split across four brands gives 2,500. ## What the benchmarks show - A leader's share varies widely. The leading brand's share runs from 37.0% in UPI apps to 91.4% in beauty e-commerce. A leader under 40% is not necessarily weak; it depends on how many brands share the set and how close the next one is. - So does concentration. Search is most concentrated in beauty e-commerce (HHI 8,428) and least in stockbroking (HHI 3,043). Two categories can have leaders with similar shares and very different races behind them. - Close races change hands. In the 4 categories where the leader is less than ten points ahead (food delivery, UPI apps, stockbroking and mobile networks), the lead has changed hands 7.0 times on average since November 2021; in the other 5, 4.0 times. A narrow lead deserves a monthly look. ## The three comparisons that make a share of search good Against an even split. With four brands in the set, an even split is 25%. A brand above its even split is taking more than its share of the category's attention; one below it is taking less. That is the fairest first benchmark for a challenger, because it ignores the leader entirely. Against your market share. This is the comparison that matters most. Les Binet's work for the IPA found that share of search tends to move before share of market, so a share of search above your market share suggests growth to come, and one below it is an early warning. The gap is the search equivalent of what planners call excess share of voice, and we explain how the two measures differ in share of search vs share of voice (https://colleviate.com/blog/share-of-search-vs-share-of-voice.html). Against last year. Direction beats level. Compare a three-month average with the same three months a year earlier, so seasonal swings cancel out. A smaller share that is rising is better news than a larger one that is slipping, and a single month means little: one news story can move a share several points for a few weeks. ## How to set your own benchmark - Define the set as your buyers see it: the brands they actually consider, not the ones you think of as rivals. - Build the monthly series in one Google Trends request, as set out in how to measure share of search (https://colleviate.com/blog/how-to-measure-share-of-search.html). - Work out the even split for the set, and note who is above it. - Put your share of search next to your best estimate of market share. If the two disagree, find out why before reporting either. - Report a three-month average against a year ago, with the competitive set alongside, every month. For categories outside the index, or a set the index does not cover, the same read is the first page of every growth audit (https://colleviate.com/blog/growth-audit-what-it-should-contain.html) we run. ## Sources - Binet presents fast, cheap, predictive Share of Search metric, IPA, 2020: https://ipa.co.uk/news/binet-presents-fast-cheap-predictive-share-of-search-metric/ - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What is a good share of search?** One above your market share and rising, measured against the brands your buyers actually compare you with. There is no universal number, because a share of search depends on how many brands share the set. **What share of search does a market leader usually have?** In the nine Indian categories in the Colleviate Share of Search Index, leaders hold between 37.0% and 91.4% of branded search. Where the leader is less than ten points ahead, the lead changes hands far more often. **Should share of search match market share?** Not exactly. Share of search tends to move before market share, so the gap between them is informative: a share of search above market share suggests growth to come, and one below it is an early warning. --- # Share of search vs share of voice: what each one tells you URL: https://colleviate.com/blog/share-of-search-vs-share-of-voice.html Topic: Share of search | Published: 2026-10-26 | Updated: 2026-10-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Share of voice is a brand's share of the category's advertising, usually measured as spend or impressions. Share of search is its share of the category's branded search. The first is an input a brand controls; the second is a response it cannot buy directly. Read together, share of voice explains a change in share of search, and share of search shows whether the voice is landing. Share of voice measures what you put into the market. Share of search measures what comes back. Reported together, they tell a board whether the money is turning into demand. ## What share of voice measures Share of voice is a brand's share of the advertising in its category. Traditionally it is share of media spend, sometimes share of impressions or ratings points, always against a named set of competitors. Its best-known use in planning is excess share of voice: the gap between a brand's share of voice and its share of the market. A long-standing finding in advertising effectiveness research is that brands whose share of voice runs ahead of their market share tend to grow, and those whose voice falls behind tend to shrink. The term has also been borrowed by social listening tools, which call a brand's share of online mentions its share of voice. That is a third measure, earned attention, and it is worth keeping separate: it moves with news and controversy as much as with anything a brand does, as reputation is a search result (https://colleviate.com/blog/reputation-is-a-search-result.html) explains. ## What share of search measures Share of search is a brand's share of the branded searches in its category, against the same named set. It records people who already had the brand in mind and went looking for it. Les Binet set it out at the IPA in 2020 and found that it tracked market share in the categories he tested, moving first. The full case is in share of search, the brand metric that leads revenue (https://colleviate.com/blog/share-of-search-leading-indicator.html). ## The difference, side by side | | Share of voice | Share of search | | What it measures | Your share of the category's advertising | Your share of the category's branded search | | Input or outcome | Input: what you put in | Outcome: what comes back | | Can you buy it? | Yes, directly | Only indirectly, through everything that makes people look for you | | Where the data comes from | Media monitoring and spend estimates, usually paid | Google Trends, free and public | | What moves it | Budgets: yours and your competitors' | Advertising, PR, product, word of mouth and news | | Best used for | Explaining why demand moved, and setting budgets | Seeing early whether demand is moving at all | ## Why share of voice is hard to measure in India Share of voice needs your competitors' spend, and in India most of that is invisible. Television and print monitoring is sold as a paid service. Digital, a growing share of most budgets, is opaque: the ad libraries Google and Meta publish show which ads ran, but not what commercial advertisers spent on them. So share of voice in India is usually an estimate built on assumptions. Share of search has no such gap, because the data is public. That is why we built and publish the Colleviate Share of Search Index (https://colleviate.com/blog/share-of-search-index-india.html) for nine Indian categories: the response side of the equation, measured the same way for every brand. ## How to read them together The two measures are most useful as a pair, because each answers the other's question. - Voice up, search up. The spend is landing. Keep it, and check that market share follows. - Voice up, search flat. The spend is not landing. Look at the creative, the targeting and whether the brand is recognisable in the ads, before adding budget. - Voice flat, search up. Something other than advertising is working: PR, product, word of mouth, a competitor's stumble. Find it, because it may be cheaper to amplify than media. - Both down. The brand is losing attention it is not paying to replace. This is the pattern to catch early. ## What to report One line for each, monthly, against the same competitive set, as three-month averages so that a single burst of spend or a single news story does not swing the reading. If you can only afford one, report share of search: it is free, it is public and it measures the outcome. Then estimate share of voice for the months where share of search moved, to explain why. What counts as a good reading of either is set out in what is a good share of search (https://colleviate.com/blog/what-is-a-good-share-of-search.html). ## Sources - Binet presents fast, cheap, predictive Share of Search metric, IPA, 2020: https://ipa.co.uk/news/binet-presents-fast-cheap-predictive-share-of-search-metric/ - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What is the difference between share of search and share of voice?** Share of voice is your share of the category's advertising, an input you control. Share of search is your share of the category's branded search, a response you cannot buy directly. Share of voice explains a change in demand; share of search shows whether demand is changing. **Which is the better leading indicator of market share?** Share of search is closer to the buyer, because it records people who already have the brand in mind. Les Binet's work for the IPA found it tracked market share and tended to move first. Share of voice is the better explanation of why it moved. **Can you measure share of voice for free?** Rarely, and not fully in India, where competitors' digital spend is not published. Share of search can be measured for free from Google Trends. --- # Festive spending in India in 2025: what the numbers show, and what they do not URL: https://colleviate.com/blog/festive-spending-india-2025.html Topic: Festive season | Published: 2026-10-21 | Updated: 2026-10-21 Publisher: Colleviate (https://colleviate.com/) **Short answer.** India’s 2025 festive season set a record on every published measure. FADA counted 52.38 lakh vehicles retailed in its 42-day festive window, up 21%; UPI handled a record 20.7 billion transactions in October; and the traders’ body CAIT estimated Diwali sales at Rs 6.05 lakh crore. GST cuts that took effect on 22 September, the first day of Navratri, drove much of it. Every number said 2025 was the biggest festive season India has recorded. They were not all measuring the same thing, and the one-off behind them will shape every comparison made this year. The 2025 festive season arrived with a tax cut attached. The reformed GST rates took effect on 22 September, the first day of Navratri, moving most goods to two main rates of 5% and 18%, with a special 40% rate for a short list of luxury and demerit goods. Small cars, two-wheelers, larger televisions and a long list of household goods became cheaper on the morning the season began. What followed was a record on every measure anyone published. The useful question for a marketer is what each of those measures actually counts. ## Vehicles: counted, and up 21% The Federation of Automobile Dealers Associations counts registrations, collated with the Ministry of Road Transport and Highways from 1,400 of India’s 1,458 regional transport offices. Over its 42-day festive window, from 22 September to 2 November 2025, retail sales reached 52,38,401 vehicles, up 21.1% on the 2024 window. Two-wheelers rose 21.8% to 40,52,503 and passenger vehicles 23.4% to 7,66,918, a festive record. FADA also recorded what happened before the cut. It described the first 21 days of September as muted, with buyers waiting for the new rates. Demand that would have arrived in early September arrived in late September and October instead. ## Payments: counted, and a record month Unified Payments Interface figures from the National Payments Corporation of India are counts of transactions. October 2025 brought 20.7 billion of them, worth Rs 27.28 lakh crore, the highest monthly volume UPI had recorded, up from 19.63 billion in September. On 18 October, the day of Dhanteras, the system crossed 750 million transactions in a single day for the first time. UPI grows every year regardless of the season, so a record month says less about festive demand than it seems to. The single-day record on the main buying day of the season says more. ## Trade: estimated, at Rs 6.05 lakh crore The Confederation of All India Traders put Diwali 2025 sales at Rs 6.05 lakh crore, Rs 5.40 lakh crore of goods and Rs 65,000 crore of services, 25% above its figure of Rs 4.25 lakh crore for 2024. It is the most quoted festive number in India and it is an estimate: a survey by CAIT’s research arm across 60 distribution centres, including state capitals and tier 2 and tier 3 cities. Read it as a direction and a mix from a body that represents traders, not as an audited total. The mix is still worth reading. CAIT attributed 12% of the trade to groceries and FMCG, 10% to gold and jewellery, 8% to electronics and electricals, and 7% each to consumer durables, ready-made garments and gift items. ## Online: modelled, and up about a quarter at the start E-commerce sale figures come from consultancies modelling platform data. Redseer Strategy Consultants estimated that online gross merchandise value grew 23% to 25% year on year across 22 and 23 September, the opening days of the sales, which coincided with the new GST rates. It credited smartphones and televisions, where GST on screens above 32 inches fell from 28% to 18%, and it had forecast growth of about 20% to 25% for the festive season as a whole. ## Three kinds of number Side by side, the season was reported in three different currencies. Registrations and payments are counted. Trade totals are surveyed estimates. Platform sales are modelled by analysts. All three pointed the same way in 2025, which is reassuring, but they should never be added together, set against each other or quoted with the same confidence. A board paper that puts a counted number beside an estimated one should say which is which. ## The calendar flatters October FADA’s own release holds the clearest warning in the data. Vehicle retail in October 2025 rose 40.5% on October 2024. Across the 42-day festive window it rose 21.1%. Same industry, same season, half the growth. The difference is the calendar. In 2025 Dussehra (2 October) and Diwali (20 October) fell in the same month. In 2024 Diwali fell on 31 October and FADA’s festive window ran to 13 November, so much of that season’s buying, and more of its registrations, landed in November. A comparison of October with October measured the festival moving, not demand growing. FADA defines its window from the first day of Navratri to 15 days after Dhanteras for exactly this reason. ## What 2026 will be compared against This year the festival moves the other way. Dussehra falls on 20 October and Diwali on 8 November 2026, so a large share of festive buying shifts out of October and into November. Expect October 2026 to look weak against October 2025 and November to look strong, for reasons that have nothing to do with anyone’s marketing. The base is also inflated. 2025 had a one-off: a tax cut on the first day of the season, after three weeks in which buyers held back. Without another one, festive growth in 2026 is measured against a year in which policy pulled demand into a few weeks. A brand that grows modestly this season may still have done well. So set festive targets on festival-aligned windows rather than calendar months, compare share against named competitors rather than raw growth, and agree both before the season starts. The method is in measuring a festive campaign when everything goes up (https://colleviate.com/blog/measuring-festive-campaigns.html), and the timing of search demand in the festive search calendar (https://colleviate.com/blog/festive-search-calendar-india.html). Qravitas has counted the company side as well: 2.5% of Indian listed company home pages carried festive messaging three weeks before Diwali 2026. ## Sources - FADA releases October’25 and 42 days festive period vehicle retail data, Federation of Automobile Dealers Associations, 7 November 2025: https://fada.in/images/press-release/1690d68d7a2e6cFADA%20releases%20October%202025%20&%2042-Days%20Festive%20Vehicle%20Retail%20Data.pdf - UPI hits all-time high of 20.7 bn transactions in October, Entrackr, reporting NPCI data, 1 November 2025: https://entrackr.com/news/upi-hits-all-time-high-of-207-bn-transactions-in-october-10612233 - Record-breaking Diwali sales of ₹5.40 lakh crore in goods, ₹65 thousand crores in services, Confederation of All India Traders, 2025: https://cait.in/record-breaking-diwali-sales-of-%E2%82%B95-40-lakh-crore-in-goods-65-thousand-crores-in-services-reflect-indias-economic-strength-and-swadeshi-spirit/ - Festive 2025: Day 0 + Day 1 ecommerce sales surge 25% with GST boost, demand led by smartphones and TVs, Redseer Strategy Consultants, 25 September 2025: https://redseer.com/articles/festive-2025-day-0-ecommerce-sales-surge-25-with-gst-boost-demand-led-by-smartphones-and-tvs/ - Next-generation GST reforms, New India Samachar, Press Information Bureau, 1 October 2025: https://newindiasamachar.pib.gov.in/WriteReadData/flipbook/2025/Oct/1st/English/files/basic-html/page21.html ## Questions **How big was India’s festive season in 2025?** By the counted measures, the largest on record. FADA reported 52,38,401 vehicles retailed in its 42-day festive window, up 21.1%, and UPI processed a record 20.7 billion transactions in October 2025. The traders’ body CAIT estimated Diwali trade at Rs 6.05 lakh crore, up 25%, from a survey of 60 distribution centres. **Why were festive sales so strong in 2025?** Mainly because reformed GST rates took effect on 22 September 2025, the first day of Navratri, cutting prices of small cars, two-wheelers, larger televisions and many household goods as the season began. FADA also credited Dussehra and Diwali falling in the same month, and rural demand. **Why is October a misleading month to compare?** Because Diwali moves by up to three weeks between years. Vehicle retail rose 40.5% in October 2025 but 21.1% across the festive window, since Diwali fell on 20 October in 2025 and on 31 October in 2024. Compare festive windows defined by the festival, not calendar months. **What should brands expect in the 2026 festive season?** Diwali falls on 8 November 2026, so buying shifts from October into November, and October will look weak against 2025. The 2025 base was also lifted by a one-off GST cut. Judge 2026 on festival-aligned windows and on share against competitors rather than raw growth. --- # Measuring a festive campaign when everything goes up URL: https://colleviate.com/blog/measuring-festive-campaigns.html Topic: Festive season | Published: 2026-10-21 | Updated: 2026-10-21 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Measure what the campaign added, not what the season delivered. Compare festive windows defined by the festival rather than calendar months, keep a holdout region or audience unexposed, track share of search and sales against named competitors, and judge spend on contribution after festive discounts. Everything rises in Diwali weeks; the question is whether you rose more than you would have without the spend. In October almost every marketing dashboard in India turns green. That is the problem. During the festive season sales go up, traffic goes up, branded search goes up and return on ad spend usually goes up too, because people who were going to buy anyway are now buying in the same few weeks the media runs. A campaign can look like a triumph and have added very little. Another can look ordinary and have held on to share that would otherwise have gone to a competitor. ## Seasonal lift is not campaign lift Separate the two before the season, in writing. Seasonal lift is what would have happened with no campaign: last year’s festive window, adjusted for the brand’s underlying trend. Campaign lift is the difference between that and what actually happened. Most festive reports present the total and imply it is the second. ## Compare festival to festival Diwali moves by up to three weeks between years, so calendar comparisons break. The vehicle trade’s 2025 figures show how badly. Retail in October 2025 was up 40.5% on October 2024, but across the festive window it was up 21.1%, because Diwali fell on 20 October in 2025 and on 31 October in 2024. Define the festive window by the festival, as the Federation of Automobile Dealers Associations does (the first day of Navratri to 15 days after Dhanteras), and compare it with the same window last year. Do the same with daily data: line up days by their distance from Diwali, not by their dates. ## Hold something back The only clean answer to what the spend caused is a comparison with people who did not see it. Keep a few matched cities, or a random slice of the audience, out of the campaign, and compare. The case for doing this is strong. When researchers compared fifteen large advertising experiments at Facebook with the observational methods advertisers usually rely on, those methods often failed to recover the experimental result, even with rich data on who saw the ads (Gordon, Zettelmeyer, Bhargava and Chapsky, Marketing Science, 2019). The bias is worst when the people most likely to buy are also the people most likely to see the ads, which describes every Diwali. A festive holdout feels expensive, because it means not advertising to some people in the best weeks of the year. It costs less than a year of decisions made on the wrong number. ## Watch share, not volume Branded search for almost every consumer brand rises in festive weeks, so a rise in yours says little. Your share of branded search against named competitors says more, because the season lifts everyone and share shows who gained. Les Binet’s work for the IPA found that share of search tends to move ahead of market share, which makes it a useful early read while the season is still running. The same logic applies to sales: share of the category in the window, from panel or marketplace data where it exists, is a better test than growth. Our Share of Search Index for India (https://colleviate.com/blog/share-of-search-index-india.html) shows the method on nine categories. ## Count the pull-forward Festive offers move purchases in time as well as creating them. Some of what a campaign sells in October would have been bought in November or December at full price. Measure the weeks after the window as well as the window itself, and expect a dip. In 2025 the effect also ran the other way: FADA reported three weeks of muted demand in September as buyers waited for GST cuts announced in advance. A price change that people know is coming moves demand to the day it arrives. ## Judge it on contribution Festive discounts, bank offers, free delivery and returns all cut into the margin on every sale, while media costs rise as every brand bids for the same audiences. A return on ad spend that looks healthy in revenue can be negative in contribution. Report what each sale contributed after discounts and the full cost of the campaign, as we set out in what ROAS tells you and what it hides (https://colleviate.com/blog/performance-marketing-roas-reality.html). ## A festive scorecard - A festive window defined by the festival, compared with the same window last year. - A holdout, by geography or audience, with the comparison agreed before launch. - Share of branded search and, where available, share of category sales against named competitors. - Contribution after discounts and the full campaign cost, not revenue over media spend. - The four weeks after the window, to see what was pulled forward. Agree all five before the first rupee is spent. Agreed afterwards, every one of them becomes a negotiation about whether the campaign worked. Check the tracking before you trust any of it: Qravitas found that 12% of company websites with analytics lose UTM campaign tags in a redirect before the visit is recorded. ## Sources - FADA releases October’25 and 42 days festive period vehicle retail data, Federation of Automobile Dealers Associations, 7 November 2025: https://fada.in/images/press-release/1690d68d7a2e6cFADA%20releases%20October%202025%20&%2042-Days%20Festive%20Vehicle%20Retail%20Data.pdf - A comparison of approaches to advertising measurement: evidence from big field experiments at Facebook, Marketing Science 38(2), 2019: https://pubsonline.informs.org/doi/10.1287/mksc.2018.1135 - Binet presents fast, cheap, predictive Share of Search metric, IPA, 2020: https://ipa.co.uk/news/binet-presents-fast-cheap-predictive-share-of-search-metric/ ## Questions **How do you measure the results of a Diwali campaign?** Compare the festive window with the same festival-aligned window last year, keep a holdout region or audience unexposed to estimate what the campaign caused, track share of search and sales against competitors, and judge the spend on contribution after discounts rather than on revenue-based ROAS. **Why can’t I compare this October with last October?** Because Diwali can fall three weeks earlier or later. In 2025 vehicle retail rose 40.5% in October but 21.1% across the festive window, since Diwali moved from 31 October to 20 October. Comparing calendar months measures the festival moving, not your marketing working. **Is a festive holdout worth the lost sales?** Usually, yes. Experiments at Facebook showed observational methods often fail to recover what advertising actually caused. A small, matched holdout in the biggest weeks of the year is the cheapest way to know what the festive budget bought before next year’s is set. --- # How to plan a festive campaign in India URL: https://colleviate.com/blog/how-to-plan-a-festive-campaign.html Topic: Festive season | Published: 2026-10-21 | Updated: 2026-10-21 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Work back from the festival date, not the calendar month. Decide whether the season is for winning new buyers or harvesting existing demand, be visible before the searching starts, and weight budget to the demand curve: half the extra searching for Diwali gifts happens in the last 7 days before Diwali, while corporate gifting runs about 17 days ahead. Set up the measurement before launch. Most festive plans are built on calendar months and launched on instinct. The festival moves every year, and the searching that comes before it runs to a timetable you can read. A festive campaign in India is planned against the busiest, most expensive and most crowded weeks of the year. The plans that work are built backwards from the festival date and timed to when people actually look for things, which our festive search calendar (https://colleviate.com/blog/festive-search-calendar-india.html) measures across five seasons. ## Work back from the festival, not the month Put the dates at the top of the plan and count backwards from them. In 2026 Dussehra falls on 20 October, Dhanteras on 6 November and Diwali on 8 November. A plan built on calendar months breaks every year, because the festival moves by up to three weeks: Diwali fell on 20 October in 2025 and on 31 October in 2024. ## Decide what the season is for Festive weeks can recruit new buyers, harvest demand that already exists, or build memory for the year ahead. Each needs different media, different creative and a different measure of success. A plan that tries to do all three usually ends up doing the middle one by default, because harvesting is what the dashboards reward. ## Be visible before the searching starts Reach that lands after people have started searching arrives while they shortlist, not before. In the calendar, a quarter of the season’s extra searching for Diwali gifts is done 16 days before Diwali and half of it 7 days before. Searches for Diwali sales are half done 13 days before, and corporate gifting runs earliest, a quarter done 34 days out. Brand-building activity belongs in the weeks before those points, not on top of them. ## Phase the budget to the demand curve Weight spend to when the category is searching. Build reach in the weeks before the curve rises, put conversion budget into the steepest part of the climb, and pull back quickly after Diwali, when the extra searching for gifts, decoration and sales is over within days. Decoration comes last among the gift, clothing and home searches, half done 5 days before Diwali. ## Make work that could only be yours Every brand runs its festive film in the same three weeks, so generic warmth disappears. Distinctive assets, an idea the brand can repeat every year, and a genuine reason to be part of the festival at all are what separate the work. We set out what gives a brand that standing in what a brand owes a festival (https://colleviate.com/blog/what-a-brand-owes-a-festival.html). ## Give the moment a name, and keep it A recurring festive moment with its own name can become the way people ask for it. In the 2021 festive sale period, 23% of searches for Flipkart’s sale used the event’s own name, Big Billion Days, and 28% of searches for Amazon’s used Great Indian Festival; the rest asked for the platform and the word sale. By 2025 the figures were 67% and 38%. Search data cannot show why the gap opened, but a name kept the same year after year is an asset that a new campaign line every October cannot build. ## Lock production, stock and approvals early Media inventory, production crews, printers, couriers and the people who approve work are all at their busiest in October. Book them before the season, agree who signs off and by when, and make sure the stock and delivery promises behind the campaign will survive the demand it creates. A festive campaign that sells out on day two has wasted the rest of its budget. ## Set up the measurement before launch Agree the festive window, the comparison with last year’s window, a holdout region or audience, and the share and contribution measures before anything goes live. Everything rises in Diwali weeks, so a result read afterwards will look like success whatever happened. The method is in measuring a festive campaign when everything goes up (https://colleviate.com/blog/measuring-festive-campaigns.html). ## Plan the days after Diwali The day after Diwali, gifting and sale searches collapse, but demand does not disappear. For many categories the weeks that follow lead into the wedding season, and vehicle dealers expected marriage-season buying to carry demand on after the 2025 festival. Decide in advance what runs then and what is switched off, so the post-festive budget is a plan rather than whatever is left. ## Sources - Colleviate Festive Search Calendar, India, Colleviate, October 2026: https://colleviate.com/blog/festive-search-calendar-india.html - FADA releases October’25 and 42 days festive period vehicle retail data, Federation of Automobile Dealers Associations, 7 November 2025: https://fada.in/images/press-release/1690d68d7a2e6cFADA%20releases%20October%202025%20&%2042-Days%20Festive%20Vehicle%20Retail%20Data.pdf ## Questions **When should a Diwali campaign start?** Before the searching does. Across five seasons, a quarter of the extra searching for Diwali gifts was done 16 days before Diwali, and a quarter of corporate gifting searches 34 days before, so brand-building reach should land in the weeks ahead of those points, with conversion spend concentrated as the searching climbs. **How should a festive budget be phased?** In line with the demand curve: reach before the curve rises, the heaviest conversion spend through the steepest part of the climb in the final fortnight, and a sharp cut after Diwali, when gifting and sale searches fall away within days. **When is Diwali in 2026?** Diwali falls on Sunday 8 November 2026 and Dussehra on 20 October, according to the central government’s list of holidays for 2026. Dhanteras, the main buying day, falls two days before Diwali, on 6 November. --- # Corporate Diwali gifts people actually keep URL: https://colleviate.com/blog/corporate-diwali-gifts.html Topic: Festive season | Published: 2026-10-21 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Give something useful, well made and quietly branded, and get it delivered before Dhanteras. Order early: across five seasons, half the extra searching for corporate gifts was done 17 days before Diwali, against 7 days for personal Diwali gifts, because made-to-order pieces take time. Keep the logo small, suit people who may not celebrate Diwali, and write the note by hand. Most corporate Diwali gifts are forgotten within a week. A few are used for years, and every use is a small reminder of who sent them. Every October, companies send something to their clients, partners and teams. Most of it disappears quickly: sweets shared around an office, a branded gadget left in a drawer, a hamper whose packaging outlives its contents. The gifts that last have a few things in common, and none of them is the size of the budget. ## A gift is a brand touchpoint A corporate gift is the company in physical form, held in someone’s hands. Its quality, restraint and usefulness are read as the company’s own. A flimsy power bank with a large logo says something about the sender, and it is rarely what the sender meant to say. Most are quiet online too: Qravitas found that 0.8% of Indian listed company home pages carried a festive greeting three weeks before Diwali 2026. ## Start earlier than feels necessary Businesses plan their gifting well ahead of everyone else. In our festive search calendar (https://colleviate.com/blog/festive-search-calendar-india.html), half the season’s extra searching for corporate gifts is done 17 days before Diwali, against 7 days for personal Diwali gifts, and a quarter of it 34 days out. Anything made to order, embroidered, printed or specially packed needs weeks rather than days, and workshops and couriers are at their busiest in October. Work back from a delivery date before Dhanteras, which falls on 6 November in 2026. ## Useful beats branded The gifts that survive are the ones that earn a place in someone’s day: a good notebook, a tote that is actually carried, a tee or hoodie someone would wear without being asked, a well made bottle or mug. Choose the object for the person, then decide how the brand appears on it, not the other way round. ## Keep the mark small The bigger the logo, the less the object is used in public, which defeats the point of putting it there. A small mark, tone on tone or placed where the owner sees it more than anyone else, gets worn and carried. It is the principle behind our own signature swag (https://colleviate.com/merch): the mark alone, small, on objects made to be used. ## Not everyone celebrates Diwali A client list or a team in India spans faiths, regions and festivals. A gift framed around the season and the relationship, rather than one ritual, travels better. Sweets and dry fruits raise dietary questions, alcohol excludes many people, and religious imagery can sit awkwardly with people who do not share it. When in doubt, send a gift for the season and a note about the year. ## The note matters more than the hamper A handwritten line that names something specific, a project, a moment, a thank-you that could only be for this person, is what gets remembered. A printed card signed by nobody is not. ## Packaging that is not straight rubbish Plastic-wrapped hampers in rigid boxes go into the bin within minutes, and the recipient notices. Pack in something reusable, or in very little. Packaging is part of the gift, as we argue in packaging as media (https://colleviate.com/blog/packaging-as-media.html). ## Logistics decide whether it lands Confirm addresses before ordering, because hybrid teams and client staff move. Deliveries to offices in the last days before Diwali meet closed reception desks and people on leave. Aim for arrival in the week before Dhanteras, and keep a short list of people who should receive theirs at home. Related. For the wider campaign, see how to plan a festive campaign (https://colleviate.com/blog/how-to-plan-a-festive-campaign.html). ## Sources - Colleviate Festive Search Calendar, India, Colleviate, October 2026: https://colleviate.com/blog/festive-search-calendar-india.html ## Questions **When should companies order corporate Diwali gifts?** Early. Across five seasons, half the extra searching for corporate gifts was done 17 days before Diwali, well ahead of personal gifting. Made-to-order pieces need several weeks, so confirm them before October’s rush and plan delivery for the week before Dhanteras. **What makes a good corporate Diwali gift?** Something useful and well made that the recipient would choose to keep: a good notebook, a tote, a tee or hoodie they would actually wear. Quality and restraint say more about the sender than the size of the hamper. **Should a corporate gift carry the company logo?** Yes, but small. A large logo stops people using the object in public; a small, tone-on-tone mark gets worn and carried, which is the point of branding it at all. --- # The festive search calendar: when India searches for Diwali, day by day URL: https://colleviate.com/blog/festive-search-calendar-india.html Topic: Festive season | Published: 2026-10-21 | Updated: 2026-10-21 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Festive search in India arrives in a fixed order. Across five seasons from 2021 to 2025, half the extra searching for corporate gifts was done 17 days before Diwali, for Diwali sales 13 days before, for Diwali gifts 7 days before and for decoration 5 days before. Dhanteras and gold coins peak on Dhanteras; greetings peak on the day. Festive demand does not arrive in one wave. It arrives in a sequence, and the sequence has barely moved in five years. Download the data (CSV) (https://colleviate.com/data/colleviate-festive-search-calendar-india.csv) Daily, five festive seasons, 2021 to 2025. Free to use with attribution. ## What the calendar shows - Business gifting is searched first. Half the season’s extra searching for corporate gifts is done 17 days before Diwali, and a quarter of it 34 days before. - Searches for Diwali sales come next, half done 13 days before Diwali. - Personal gifts and outfits crowd the final stretch: half done 7 and 8 days before Diwali. - Decoration comes last among the gift, clothing and home searches, half done 5 days before. - Dhanteras and gold coins peak 2 days before Diwali, on Dhanteras itself; greetings peak on Diwali itself. - Corporate gifts, sales, personal gifts and decoration reached their halfway points in that order in every one of the five seasons. Each search starts from its own quiet level in August, rises as the festival approaches and falls back within days of Diwali. What differs is when it rises. The calendar below shows, for each search, when a quarter, half and three quarters of the season’s extra searching had happened, as the median of five seasons. Quarter to three quarters of the extra searchingHalf donePeak day Days relative to Diwali (Lakshmi Puja). Median of the 2021 to 2025 festive seasons, India. Source: Google Trends; Colleviate analysis. ## Business gifting starts first Companies plan ahead of everyone else. Searches for corporate gifts run all year, and the festive rise on top of them starts early: a quarter of it is done 34 days before Diwali and half 17 days before, against 7 days for personal Diwali gifts. Made-to-order gifts need that lead time, which is why the guide to corporate Diwali gifts people keep (https://colleviate.com/blog/corporate-diwali-gifts.html) starts with the calendar. ## Deals before gifts Searches for Diwali sales are half done 13 days before Diwali and peak a median 11 days out, well ahead of searches for gifts. People look for the deal before they decide what to buy with it. ## The final stretch belongs to gifts, clothes and decoration Personal gifting builds for weeks but is concentrated late: half the extra searching for Diwali gifts happens in the last 7 days, and it peaks 2 days before Diwali. Outfits follow the same shape, half done 8 days before and peaking on Diwali itself, and decoration comes last, half done 5 days before. Corporate giftsDiwali saleDiwali giftsDiwali decoration A typical season: seven-day centred average of daily search interest, each season scaled to its own peak, median of 2021 to 2025, India. Source: Google Trends; Colleviate analysis. ## Dhanteras and gold: decided in days Searches for Dhanteras and for gold coins rise late and sharply, and both peaked 2 days before Diwali, on Dhanteras, in every one of the five seasons. Dhanteras is the traditional day for buying gold, silver, utensils and vehicles, and for a jeweller or a vehicle brand the season is decided in a few days rather than a few weeks. ## Greetings arrive on the day Searches for Diwali wishes peak on Diwali itself. Content made to be shared, greetings, cards and messages, is looked for on the day itself, when shopping searches have already fallen away. ## The calendar for 2026 Diwali falls on 8 November 2026. Applying the five-season medians to that date gives the dates below. They are a guide to timing, not a forecast of any brand’s demand. On the medians, a quarter of this season’s corporate gifting searches were done by 5 October, while personal gifting reaches its quarter mark around 23 October. | Search | A quarter done | Half done | Peak day | | Corporate gifts | 5 October | 22 October | 1 November | | Diwali sale | 18 October | 26 October | 28 October | | Diwali outfit | 24 October | 31 October | 8 November | | Diwali gifts | 23 October | 1 November | 6 November | | Diwali decoration | 28 October | 3 November | 5 November | | Gold coin | 29 October | 5 November | 6 November | | Dhanteras | 5 November | 6 November | 6 November | | Diwali wishes | 7 November | 8 November | 8 November | ## Sale names: how much searching the event name captures The two largest online festive sales have names of their own. In the sale period of each season, from 45 days before Diwali to 5 days after, we compared searches for each event name with plain searches for the platform and the word sale. Flipkart searches using Big Billion DaysAmazon searches using Great Indian Festival Share of each platform’s festive sale searches that used the event name rather than the platform name with “sale”. India. Source: Google Trends; Colleviate analysis. A recurring festive moment with its own name can become the way people ask for it. In the 2021 festive sale period, 23% of searches for Flipkart’s sale used the event’s own name, Big Billion Days, and 28% of searches for Amazon’s used Great Indian Festival; the rest asked for the platform and the word sale. By 2025 the figures were 67% and 38%. Search data cannot show why the gap opened, but a name kept the same year after year is an asset that a new campaign line every October cannot build. Great Indian Festival is also an ordinary phrase, so some of its searches will not be about the sale at all, which if anything flatters Amazon’s figure. Data table | Season | Flipkart: searches using Big Billion Days | Amazon: searches using Great Indian Festival | | 2021 | 23% | 28% | | 2022 | 33% | 31% | | 2023 | 50% | 37% | | 2024 | 59% | 38% | | 2025 | 67% | 38% | ## The full table Days relative to Diwali, median of five seasons; a negative number is days before. The last column shows the earliest and latest halfway point across the five seasons. | Search | A quarter done | Half done | Three quarters done | Peak day | Half done, earliest to latest season | | Corporate gifts | -34 | -17 | -9 | -7 | -24 to -15 | | Diwali sale | -21 | -13 | -6 | -11 | -14 to -10 | | Diwali outfit | -15 | -8 | -2 | 0 | -9 to -5 | | Diwali gifts | -16 | -7 | -3 | -2 | -8 to -6 | | Diwali decoration | -11 | -5 | -2 | -3 | -5 to -4 | | Gold coin | -10 | -3 | -2 | -2 | -5 to -2 | | Dhanteras | -3 | -2 | -2 | -2 | -2 to -2 | | Diwali wishes | -1 | 0 | 0 | 0 | 0 to 0 | ## How the calendar is built For each search term and each festive season from 2021 to 2025, we requested daily Google Trends data for India over a 91-day window, from 75 days before Diwali to 15 days after it, one term per request, so every day in every window sits at a known distance from Diwali. Diwali is Lakshmi Puja as listed in the central government’s holiday calendar for that year. The data was retrieved on 19 October 2026. A term’s baseline is the median of the first 15 days of its window, well before the festival. Its festive lift on any day is its search interest above that baseline. Between 60 days before Diwali and 10 days after, we find the days by which a quarter, half and three quarters of the season’s lift had happened, and the day of the peak. The calendar reports the median of the five seasons for each. For the sale names, each season is one request containing Big Billion Days, Great Indian Festival, Flipkart sale and Amazon sale, so all four share one scale. The figures sum each term over the sale period, 45 days before Diwali to 5 days after, which leaves out the Independence Day sales in August. ## What it does not tell you Google Trends reports relative interest, not volumes, and each season of each term here has its own scale, so the calendar reads timing within a season, not growth between seasons. The terms are in English; a great deal of festive searching in India happens in Hindi and other languages, in Roman script and in their own, and on marketplaces, YouTube and social platforms rather than search engines. Search is also not purchase: the calendar shows when people look, and buying usually follows within days. The 2021 season took place as pandemic restrictions eased, though its timings sit close to the other four. ## Use and citation The data is free to use, quote and republish under a Creative Commons Attribution 4.0 licence. Please cite it as: Colleviate Festive Search Calendar, India, October 2026, colleviate.com, with a link to this page. Source data: Google Trends. The same read can be built for a brand’s own category and competitors; it is part of the share of search work (https://colleviate.com/blog/share-of-search-index-india.html) we publish for India. ## Sources - Google Trends: understanding the data, Google News Initiative, undated: https://newsinitiative.withgoogle.com/resources/lessons/google-trends-understanding-the-data/ ## Questions **When do people in India start searching for Diwali gifts?** The rise starts weeks out, but most of it is late: a quarter of the season’s extra searching for Diwali gifts is done 16 days before Diwali and half of it 7 days before, as the median of five seasons from 2021 to 2025. **When do businesses search for corporate Diwali gifts?** Earlier than consumers. Half the season’s extra searching for corporate gifts is done 17 days before Diwali and a quarter 34 days before. **Where does the data come from?** Google Trends, for India, at daily resolution, one search term per request, in a 91-day window around Diwali for each season from 2021 to 2025. Timings are measured relative to Diwali, so seasons can be compared although the date moves. **Can I use the data?** Yes. It is published under a Creative Commons Attribution 4.0 licence: use, quote and republish it with credit to the Colleviate Festive Search Calendar and a link to this page. --- # Share of search in India: nine categories, five years of data URL: https://colleviate.com/blog/share-of-search-index-india.html Topic: Share of search | Published: 2026-10-14 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Across nine Indian categories from November 2021 to September 2026, Ather rose to 62.2% of electric two-wheeler brand search as Ola Electric fell to 14.7%, Groww overtook Zerodha in stockbroking, Blinkit leads quick commerce with 56.4%, and Birla Opus reached 11.0% of paints from nothing. The monthly data is free to download. Share of search moves before market share and can be read entirely from public data. So we read it for nine Indian categories, every month for five years, and published the numbers. Download the data (CSV) (https://colleviate.com/data/colleviate-share-of-search-india.csv) Monthly, November 2021 to September 2026. Free to use with attribution. ## What the index shows - Electric two-wheelers moved most: Ather rose to 62.2% of the set's branded search as Ola Electric fell from 48.7% to 14.7%. - In stockbroking, Groww (40.1%) has led Zerodha (30.2%) every month since September 2025. - Blinkit holds 56.4% of quick-commerce brand search; BigBasket fell from 44.5% to 4.9%. - Birla Opus went from no measurable search interest to 11.0% of decorative paints, ahead of two established brands. - PhonePe (37.0%) has overtaken Paytm (35.6%), which held 67.8% at the start. - Swiggy and Zomato are level at about half each, and the lead has changed hands repeatedly since November 2024. Share of search is the proportion of branded search interest a brand captures against its named competitors. It tends to move before market share, which is why it is worth reading early; we explain the measure in share of search, the brand metric that leads revenue (https://colleviate.com/blog/share-of-search-leading-indicator.html), and how to build it yourself in how to measure share of search (https://colleviate.com/blog/how-to-measure-share-of-search.html). ## Quick commerce: Blinkit pulled away BigBasket held 44.5% of the set's branded search when the series begins and 4.9% now. Blinkit rose from 13.1% to 56.4% (the brand took the Blinkit name in December 2021, so its line starts from zero) and has led every month since February 2025, after Zepto held the lead from October 2024 to January 2025. Zepto sits at 28.0% and Swiggy Instamart at 10.7%. Read BigBasket with care: it is a scheduled grocery service as well as a quick-commerce one, so part of its fall is likely the category changing shape rather than a brand losing. BlinkitZeptoSwiggy InstamartBigBasketShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | Blinkit | 13.1% | 50.8% | 56.4% | +5.6 pts | | Zepto | 21.1% | 31.6% | 28.0% | -3.6 pts | | Swiggy Instamart | 21.4% | 12.3% | 10.7% | -1.6 pts | | BigBasket | 44.5% | 5.2% | 4.9% | -0.3 pts | The full quick commerce read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-quick-commerce-india.html) ## Food delivery (company brands): level, and swapping the lead Zomato held 61.5% at the start. The two are now level, 50.9 to 49.1 in Swiggy's favour, and the lead has changed hands repeatedly since November 2024. Searches for Swiggy include people looking for Instamart, so this is interest in the company brands rather than in food delivery alone. SwiggyZomatoShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | Swiggy | 38.5% | 49.3% | 50.9% | +1.6 pts | | Zomato | 61.5% | 50.7% | 49.1% | -1.6 pts | The full food delivery read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-food-delivery-india.html) ## UPI payment apps: Paytm lost half its share; PhonePe took the lead Paytm fell from 67.8% to 35.6%. PhonePe rose from 13.6% to 37.0%, first overtook Paytm in October 2025 and has led since May 2026. Paytm's own peak, 75.6% in February 2024, was the month the RBI restricted Paytm Payments Bank: attention, not demand. Google Pay holds 21.5% and CRED 5.9%. Search share is not payment share: Paytm's searches include its share price and its banking news, and an app used daily from the home screen rarely needs searching for. PhonePeGoogle PayPaytmCREDShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | PhonePe | 13.6% | 36.5% | 37.0% | +0.5 pts | | Google Pay | 15.8% | 17.8% | 21.5% | +3.7 pts | | Paytm | 67.8% | 39.2% | 35.6% | -3.6 pts | | CRED | 2.7% | 6.5% | 5.9% | -0.6 pts | The full upi payment apps read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-upi-payments-india.html) ## Stockbroking: Groww overtook Zerodha Zerodha's share fell from 61.2% to 30.2%, while Groww rose from 14.4% to 40.1% and Angel One from 3.2% to 21.2%. Groww first led in January 2024 and has led every month since September 2025; its spike to 52.1% in November 2025 was the month of its stock market listing. Dhan is left out of the share: the word also means wealth in Hindi and appears inside other names, which would overstate it. ZerodhaGrowwUpstoxAngel OneShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | Zerodha | 61.2% | 33.5% | 30.2% | -3.3 pts | | Groww | 14.4% | 33.1% | 40.1% | +7.0 pts | | Upstox | 21.2% | 10.2% | 8.6% | -1.6 pts | | Angel One | 3.2% | 23.2% | 21.2% | -2.0 pts | The full stockbroking read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-stockbroking-india.html) ## Electric two-wheelers: the largest shift in the index Ola Electric held 48.7% of the set's branded search at the start and 14.7% now. Ather rose from 35.2% to 62.2% and has led every month since September 2025. Ola Electric's peak search interest came in August 2024, the month of its stock market listing, a reminder that attention and buying intent are not the same thing. TVS iQube (13.5%) and Bajaj Chetak (9.5%) have held steadier. Ola ElectricAtherTVS iQubeBajaj ChetakShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | Ola Electric | 48.7% | 39.1% | 14.7% | -24.4 pts | | Ather | 35.2% | 38.1% | 62.2% | +24.1 pts | | TVS iQube | 7.6% | 13.7% | 13.5% | -0.2 pts | | Bajaj Chetak | 8.5% | 9.1% | 9.5% | +0.4 pts | The full electric two-wheelers read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-electric-two-wheelers-india.html) ## Decorative paints: a new entrant reached double digits Asian Paints still holds 72.1%, down from 77.6%. Birla Opus first registered search interest in September 2023 and now holds 11.0%, ahead of Nerolac (7.6%) and Berger Paints (7.3%). Indigo Paints holds 2.0% and is shown in the table only, to keep the chart to four lines. Asian PaintsBerger PaintsNerolacBirla OpusShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | Asian Paints | 77.6% | 75.6% | 72.1% | -3.5 pts | | Berger Paints | 12.1% | 7.3% | 7.3% | +0.0 pts | | Nerolac | 6.4% | 7.1% | 7.6% | +0.5 pts | | Birla Opus | 0.0% | 8.3% | 11.0% | +2.7 pts | | Indigo Paints | 3.8% | 1.7% | 2.0% | +0.3 pts | The full decorative paints read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-paints-india.html) ## Domestic airlines: attention is not demand IndiGo leads with 55.6%, close to where it started. Air India rose from 33.5% to 37.8%. Its only month in the lead was June 2025, the month of the Ahmedabad crash, which is why every spike in share of search needs a cause attached before anyone reads it as demand. SpiceJet fell from 12.2% to 3.6%; Akasa Air grew from almost nothing to 3.0%. IndiGoAir IndiaAkasa AirSpiceJetShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | IndiGo | 54.0% | 54.9% | 55.6% | +0.7 pts | | Air India | 33.5% | 37.6% | 37.8% | +0.2 pts | | Akasa Air | 0.2% | 3.1% | 3.0% | -0.1 pts | | SpiceJet | 12.2% | 4.3% | 3.6% | -0.7 pts | The full domestic airlines read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-airlines-india.html) ## Mobile networks (prepaid recharge): measured on recharges, not names Bare searches for Jio or Airtel also match streaming, banking and finance brands, so this category uses prepaid recharge searches instead. Jio leads with 44.4%, Airtel follows with 34.9%. BSNL briefly led in July 2024, the month the private operators raised tariffs, and has since fallen to 10.2%. Vi holds 10.6%. JioAirtelViBSNLShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | Jio | 37.6% | 43.2% | 44.4% | +1.2 pts | | Airtel | 37.3% | 32.6% | 34.9% | +2.3 pts | | Vi | 8.1% | 9.7% | 10.6% | +0.9 pts | | BSNL | 17.0% | 14.5% | 10.2% | -4.3 pts | The full mobile networks read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-telecom-india.html) ## Beauty e-commerce: one brand owns the search Nykaa holds 91.4% of branded search against Purplle's 8.6%. Tira Beauty, Reliance's beauty platform, is left out because its volume sits below the level at which Google Trends reports a value. NykaaPurplleShare of branded search within the set, by month, India, November 2021 to September 2026. Source: Google Trends; Colleviate analysis.Data table | Brand | Start of series | A year ago | Last three months | Change in a year | | Nykaa | 95.5% | 91.1% | 91.4% | +0.3 pts | | Purplle | 4.5% | 8.9% | 8.6% | -0.3 pts | The full beauty e-commerce read: who has led and for how long, the months that moved it, and the monthly data (https://colleviate.com/blog/share-of-search-beauty-ecommerce-india.html) ## How the index is built Each category is one Google Trends request for India over five years at weekly resolution, with every brand in the set compared inside that request, so all of them share one scale. A brand's monthly share is its summed weekly interest divided by the summed interest of the whole set. The partial first and last months of the window are dropped. The data was retrieved on 2026-10-14. The search terms are the names people type. Where a bare name collides with something else, the term is qualified or the brand is left out, and the reason is published below. Google Trends reports relative interest on a 0 to 100 scale and rounds very small values to zero, so brands with little search volume read as zero rather than as small. | Category | Search terms | Left out of the share | | Quick commerce | Blinkit, Zepto, Instamart, BigBasket | None | | Food delivery (company brands) | Swiggy, Zomato | None | | UPI payment apps | PhonePe, Google Pay, Paytm, CRED | None | | Stockbroking | Zerodha, Groww, Upstox, Angel One | Dhan: also the Hindi word for wealth, and part of other names such as Jan Dhan | | Electric two-wheelers | Ola Electric, Ather, TVS iQube, Bajaj Chetak | None | | Decorative paints | Asian Paints, Berger Paints, Nerolac, Birla Opus, Indigo Paints | None | | Domestic airlines | IndiGo, Air India, Akasa Air, SpiceJet | None | | Mobile networks (prepaid recharge) | Jio recharge, Airtel recharge, Vi recharge, BSNL recharge | None | | Beauty e-commerce | Nykaa, Purplle | Tira Beauty: below the threshold at which Google Trends reports a value | ## What it does not tell you Share of search measures attention, not sales. A spike can be news, a share price, a controversy or a launch, so every movement needs a cause before it is read as demand. It also only counts search engines: in India a growing share of discovery starts on marketplaces, YouTube and quick-commerce apps, which we cover in share of search in the Indian market (https://colleviate.com/blog/share-of-search-in-india.html). Read it as a leading question about a market, not a verdict on it. ## Use and citation The data is free to use, quote and republish under a Creative Commons Attribution 4.0 licence. Please cite it as: Colleviate Share of Search Index, India, October 2026, colleviate.com, with a link to this page. Source data: Google Trends. If your category is not here, the same read can be built for any competitive set, and it is the first page of every growth audit (https://colleviate.com/blog/growth-audit-what-it-should-contain.html) we run. Related. What we would do with the readings: window shopping, the brands we admire and the campaigns we would pitch (https://colleviate.com/blog/window-shopping.html). ## Questions **What is the Colleviate Share of Search Index?** A monthly measure of each brand's share of branded search interest against its named competitors, for nine Indian consumer categories from November 2021 to September 2026, built from Google Trends and published with the data. **Where does the data come from?** Google Trends, for India, at weekly resolution over five years. Each category is a single request with every brand in it, so the brands share one scale and their shares can be compared directly. **Does share of search predict market share?** It tends to lead it, because people search during consideration, before they buy. Research popularised by Les Binet and the IPA found that changes in share of search tend to precede changes in market share by several months. It measures attention, so a spike needs a cause before it is read as demand. **Can I use the data?** Yes. It is published under a Creative Commons Attribution 4.0 licence: use, quote and republish it with credit to the Colleviate Share of Search Index and a link to this page. --- # How to test Meta ad creative properly URL: https://colleviate.com/blog/meta-ads-creative-testing.html Topic: Measurement | Published: 2026-10-08 | Updated: 2026-10-08 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Test three to five genuinely different concepts at once, not variations of one. Give each enough budget to reach roughly fifty conversions or two weeks, whichever comes first, hold audience and placement constant, and judge on cost per outcome rather than click-through rate. Most accounts are not testing. They are running several versions of the same idea and calling the least bad one a winner. Creative is the biggest lever on Meta and the one most teams treat casually. The platform will optimise delivery within what you give it. It cannot invent an idea you did not supply. ## Test concepts, not variations A different headline on the same image is not a test. It is a tweak, and it will produce a result inside the noise. Put three to five genuinely different propositions in market: a demonstration, a testimonial, a price-led cut, a problem-led cut, something odd. Hold everything else constant. Same audience, same placements, same objective, same landing page. If two things change, you have learned nothing you can carry forward. ## Give it enough to decide Each concept needs roughly fifty conversions or two weeks before the number means anything. Below that you are reading variance. The common failure is killing a concept on day three because the first hundred impressions went badly, which is how accounts end up with one safe asset and a rising cost per acquisition. ## Judge on the right metric Click-through rate tells you the thumbnail worked. Cost per qualified outcome tells you the idea worked. Those come apart more often than people expect: the asset with the best click-through is frequently the one making a promise the page cannot keep. ## What to do with the winner Scale it, then immediately start building its replacement. On an active account the main asset degrades within four to six weeks: frequency climbs, click-through falls, cost per click holds. That is fatigue, and the fix is new work rather than new audiences. The fingerprints of each cause are in why your ads stopped working (https://colleviate.com/blog/why-your-ads-stopped-working.html). ## Tests that waste money Button colour. Minor copy edits. Audience splits when the creative is the constraint. Anything run for three days. And testing during a sale period, when the offer is doing the work and the creative result will not repeat. ## Questions **How many ad creatives should we test at once?** Three to five genuinely different concepts. More than that and none of them gets enough budget to produce a readable result. **How long should a creative test run?** Until each concept has roughly fifty conversions, or two weeks. Shorter than that and you are making decisions from variance. --- # Google Ads account structure that still works URL: https://colleviate.com/blog/google-ads-account-structure.html Topic: Measurement | Published: 2026-10-08 | Updated: 2026-10-08 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Split campaigns only where you need separate budgets, separate targets or separate reporting. Within that, keep themes tight enough that one ad can answer every query in the group, run broad match only with solid conversion data and an active negative list, and review the search terms report weekly. The old advice was to build hundreds of tiny ad groups. Automation has made most of that structure pointless, and some of it actively harmful. Structure used to be how you controlled Google Ads. Bidding automation now does most of that job, and an account carved into two hundred ad groups starves each of the data the system needs to work. ## Split for a reason, not for tidiness There are three good reasons to separate campaigns: different budgets, different targets, different reporting needs. Geography when budgets differ by market. Brand versus non-brand, always, because mixing them hides the truth about both. Product lines where the economics genuinely differ. Everything else can sit together. Consolidation gives the bidding more signal, and signal is what the system runs on. ## Match types, practically Exact for the terms you know convert. Phrase for the controlled middle. Broad only when the account has steady conversion volume and a maintained negative list, because broad without either is a budget donation to adjacent intent. ## Negative keywords are the real work Read the search terms report weekly. Add negatives at the right level, keep a shared list for the obvious category exclusions, and watch for the slow drift where a campaign starts matching informational queries that will never buy. Most wasted spend in a mature account is here, not in bidding. ## Legacy habits to drop - Single keyword ad groups at scale. They fragment data for a control you no longer have. - Duplicating campaigns per match type. It splits volume and teaches the system nothing. - Bidding hard on your own brand at saturation without testing what organic absorbs. - Leaving Performance Max unsegmented and unexamined, then wondering where the money went. ## What actually moves the number The landing page, usually. A well-structured account pointing at a weak page is an expensive way to test copy, which is the argument in landing pages that convert (https://colleviate.com/blog/landing-pages-that-convert.html). ## Questions **How many campaigns should a Google Ads account have?** As few as the budgets, targets and reporting needs allow. Consolidation gives automated bidding more signal; fragmentation starves it. **Is broad match safe now?** With steady conversion data and an actively maintained negative list, yes. Without either, it spends against adjacent intent that will not convert. --- # The ad account audit checklist URL: https://colleviate.com/blog/ad-account-audit-checklist.html Topic: Measurement | Published: 2026-10-07 | Updated: 2026-10-07 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Audit in this order: conversion tracking first, because everything downstream depends on it, then wasted spend, then structure, then creative supply, then the reporting itself. Most accounts have a tracking problem that invalidates the rest of the analysis. Every agency transition involves an audit, and most of them start in the wrong place. Start with whether the numbers are real. Run this before you take over an account, before you brief a new agency, and once a year on your own account whoever is running it. ## 1. Tracking, before anything else Which conversions are counted, how they are defined, and whether duplicates exist. Compare platform-reported conversions against the finance system for one month. A twenty percent gap is normal. A gap nobody can explain means the rest of the audit is guesswork. The method is in a GA4 setup that survives scrutiny (https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html). ## 2. Waste Search terms with spend and no conversions over ninety days. Placements on display that nobody would choose deliberately. Audiences overlapping so you bid against yourself. Geography spending outside the markets you serve. This is usually where the first ten percent of savings sits. ## 3. Structure Brand separated from non-brand. Budgets matched to intent. Campaign count justified. Automated campaign types examined rather than trusted. ## 4. Creative supply How many distinct concepts ran in the last quarter, and how often assets were refreshed. An account running the same three assets for six months is not being managed, it is being maintained. ## 5. The reporting Ask to see the last three monthly reports. If every one opens with a win and none names something that failed, you are reading advocacy rather than analysis. That tells you more about the relationship than any metric in the account. ## What a good audit produces Three numbers you can act on, a list of things to stop, and a written statement of what the account would need to look like to be judged fairly in ninety days. Not forty slides. ## Questions **What should a paid media audit cover first?** Conversion tracking. If the measurement is wrong, every conclusion drawn from the account is wrong with it. **How often should we audit our own ad accounts?** Annually at minimum, and always at an agency transition or a significant budget change. --- # How to measure share of search URL: https://colleviate.com/blog/how-to-measure-share-of-search.html Topic: Share of search | Published: 2026-10-06 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Share of search is your brand's search volume as a proportion of the total for your competitive set. To measure it, choose four to eight brands a buyer genuinely considers alongside you, pick the branded terms, pull the data from a source such as Google Trends, then normalise and smooth it. Most of the difficulty is in the setup. The measure is free, takes an afternoon to set up, and moves months before revenue does. Most of the difficulty is in the setup, not the maths. Share of search is your brand’s search volume as a proportion of the total for your competitive set. That is the entire calculation. Everything that makes it useful or useless happens before you divide. ## Pick the competitive set properly Not the companies you dislike. The companies a buyer genuinely considers alongside you. Four to eight names is workable; beyond that the series gets noisy and the chart becomes decorative. Two tests. Would a customer shortlist them in the same conversation as you? Do they sell to the same buyer with a similar price and model? If the answer to either is no, leave them out and note why, because somebody will ask next quarter. ## Choose the terms Use brand terms, not category terms. “Nike” rather than “running shoes”. Brand queries are the demand that already belongs to someone; category queries are demand nobody owns yet, and mixing them produces a number that means neither thing. The two are worth tracking separately, which we set out in branded versus category search (https://colleviate.com/blog/branded-versus-category-search.html). Watch for collisions. Common-word brand names pick up unrelated searches, and a brand that shares a name with a film, a place or a person will look far healthier than it is. Where a name is ambiguous, use the brand plus a category word consistently across the whole set. ## Pull the data Google Trends is free and compares up to five terms at once, which is why the five-name limit tends to shape the set. It gives relative interest rather than absolute volume, which is fine, because share is a ratio anyway. For absolute numbers, a keyword tool or Search Console for your own brand will do. Pull at least three years. Anything shorter and you cannot tell a trend from a season. ## Normalise and smooth Convert each brand to a percentage of the set total for each period, so the series is comparable even when overall category demand moves. Then apply a rolling average, usually four or twelve weeks, because weekly data in most categories is mostly noise and reading noise as signal is the fastest way to lose faith in the measure. ## What to look at - Your floor, not your peaks. Campaigns create spikes. A rising floor between spikes is the thing that compounds. - Movement relative to the set. Everyone falling together is a category event, not your problem. - The gap between you and the leader, tracked over years rather than months. ## The mistakes that waste the effort Changing the competitive set between reports, which makes the series meaningless. Reading a single month. Including a competitor whose brand name is also a common word. Reporting the index value rather than the share. And presenting it as a performance metric when it is a leading indicator: it tells you where demand is heading, not what happened last month. Where it belongs in reporting is covered in the dashboard that gets read (https://colleviate.com/blog/marketing-dashboard-that-gets-read.html). Reading a company from public signals rather than its own data has a name, outside-in analysis , and Qravitas has written up what it involves. ## What does Google Trends actually give you? Not search volumes. Google Trends returns relative interest on a scale of 0 to 100, where 100 is the highest point in the window you asked for, and it is calculated from a sample of searches, so the same query can return slightly different numbers on different days. Terms with very little volume come back as zero, which means too few searches to report, not none at all. The practical consequence is the most important rule of the method: put every brand in the competitive set into the same request. Brands compared inside one request share one scale, so their ratio is a true share. Brands pulled in separate requests are each scaled to their own peak, and dividing one by another produces a number that means nothing. ## Which search terms break the measure? Building our share of search index for India (https://colleviate.com/blog/share-of-search-index-india.html) surfaced the failures that matter most, and they are all about names rather than maths: - Common words. Dhan is a stockbroking brand and also the Hindi word for wealth, inside names such as Jan Dhan. Its raw share looked like a third of the category; we left it out rather than publish an inflated number. - Names inside other names. A broad match on Mahindra also counts searches for Kotak Mahindra Bank, so maker names were the wrong unit for cars. - Brands that share a word with other businesses. Bare searches for Jio and Airtel include streaming, banking and finance products, so telecom is measured on recharge searches instead. - How people actually type a name. Searching for Swiggy Instamart undercounts a service most people call Instamart. - Rebrands. Blinkit took that name in December 2021, so its series starts from zero. A rename is a break in the data, not a surge. The test before trusting any term is to read the queries Google Trends lists as related to it. If they are about something else, the term is measuring something else. ## A worked example For quick commerce we put Blinkit, Zepto, Instamart and BigBasket into one request for India over five years, summed each brand's weekly values by month, and divided by the total for the set. Over the last three months of the series, Blinkit held 56.4 percent of the set's branded search, Zepto 28.0, Swiggy Instamart 10.7 and BigBasket 4.9, against 44.5 percent for BigBasket at the start. The full monthly series is in the downloadable data (https://colleviate.com/data/colleviate-share-of-search-india.csv). ## Sources - Google Trends: understanding the data, Google News Initiative, undated: https://newsinitiative.withgoogle.com/resources/lessons/google-trends-understanding-the-data/ - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What tools do I need to measure share of search?** Google Trends is enough to start and costs nothing. A keyword tool helps if you want absolute volumes rather than relative interest, and Search Console covers your own brand. **How many competitors should be in the set?** Four to eight, chosen because a buyer would genuinely consider them alongside you. Google Trends compares five terms at a time, which is a practical constraint. **How often should share of search be reported?** Monthly, on a rolling average, read quarterly. Weekly readings in most categories are noise. --- # When retargeting stops being efficient URL: https://colleviate.com/blog/retargeting-limits.html Topic: Measurement | Published: 2026-10-06 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Retargeting reports well because it reaches people already intending to buy. Test it with a holdout: suppress retargeting for a random share of your audience for a month and compare conversion rates. In most accounts a large share of the credited revenue would have arrived anyway. It is the best-performing line in most media reports and the least incremental. Both things are true at once, which is why the argument never ends. Retargeting shows people an ad for something they already looked at. Of course it converts. The question is never whether it converts, it is how many of those conversions needed the ad at all. ## Why the reported number flatters Attribution credits the last touch before purchase, and retargeting is engineered to be the last touch. Somebody who was coming back anyway sees your ad on the way, clicks it, and the channel claims the sale. Reported return is excellent. Incremental return can be close to zero. ## The test that settles it Hold out a random share of your retargeting audience for four weeks. Compare conversion rates between the suppressed group and the targeted group. The difference is the incremental effect. It is the cheapest high-value test in paid media and almost nobody runs it, because the answer is often uncomfortable. ## The frequency problem Past a certain point retargeting stops persuading and starts irritating. There is no universal number, but the signal is easy to read: rising frequency with falling click-through and no lift in conversion rate. Cap it, exclude recent purchasers, and exclude people who have seen the asset more than a handful of times. ## Sizing it honestly Retargeting should be a small, capped share of paid budget: enough to catch genuine abandonment, not enough to re-bill your existing demand. Where the real work sits is upstream, in creating demand that does not yet exist, as argued in brand versus performance (https://colleviate.com/blog/brand-versus-performance-split.html), and downstream in owned channels, which cost nothing per message: see email and lifecycle marketing (https://colleviate.com/blog/email-and-lifecycle-marketing.html). Related. The same question asked of brand search ads: what eBay learned when it switched off its brand ads (https://colleviate.com/blog/what-ebay-learned-by-switching-off-brand-ads.html). ## Questions **Is retargeting worth it?** In moderation, for genuine abandonment. Test incrementality with a holdout before scaling it, because reported returns overstate the effect substantially. **What frequency cap should we set on retargeting?** There is no universal figure. Watch for rising frequency with falling click-through and no conversion lift, and cap below that point. --- # The technical SEO checklist that matters URL: https://colleviate.com/blog/technical-seo-checklist.html Topic: Method | Published: 2026-10-05 | Updated: 2026-11-03 Publisher: Colleviate (https://colleviate.com/) **Short answer.** In order of impact: make sure pages can be crawled and indexed, that canonicals point where you intend, that the site is fast on a mid-range phone, that structured data matches the visible page, and that internal links connect related pages. Most other technical findings are noise. Technical SEO audits routinely return four hundred issues. Perhaps eight of them change anything. The purpose of technical work is to remove reasons a search engine might not use your page. It is finite, it has an end, and past that end more technical work returns nothing. ## Crawl and index Can the important pages be reached, and are they actually indexed? Check the index coverage report, not the sitemap. A page in your sitemap and not in the index is telling you something: usually that it is thin, duplicated, or orphaned from internal links (https://colleviate.com/blog/internal-linking-for-topic-clusters.html). ## Canonicals Every page should declare the URL you want indexed, and that URL should be the one in the sitemap and the one you link to internally. Mismatches between those three is the most common cause of pages quietly not ranking. ## Speed, on a real phone Test on a mid-range Android on mobile data, not on your laptop. Largest contentful paint is usually an image or a font. Interaction delay is usually third-party script. Both are fixable without a rebuild, and both matter more for conversion than for ranking. ## Structured data that matches the page Mark up what is visibly on the page and nothing else. Invented ratings and prices get manual actions, and the temptation is real because the rich result is attractive. Which markup is worth adding is covered in schema markup that earns results (https://colleviate.com/blog/schema-markup-that-earns-results.html). ## Internal linking The cheapest ranking lever most sites ignore. Every new page should link to and from the pages it belongs with, so the cluster is reachable and the relationship is legible. ## Safe to ignore Most crawler warnings about meta keyword tags, minor heading order on decorative elements, trailing slash consistency once redirects are in place, and anything described as a score out of a hundred by a tool that is also selling you something. Related. Once pages are crawlable, connect them: internal linking for topic clusters (https://colleviate.com/blog/internal-linking-for-topic-clusters.html) includes an audit for pages nothing links to. Related. When rankings move, read Search Console diagnostics (https://colleviate.com/blog/search-console-diagnostics.html); for shops, ecommerce SEO in India (https://colleviate.com/blog/ecommerce-seo-india.html). Related. What these checks looked like on a real, new site: our own Search Console, first ninety days (https://colleviate.com/blog/our-own-search-console-first-ninety-days.html). ## Questions **How often should a technical SEO audit be run?** Once properly, then after any migration or template change. Technical work is finite; recurring monthly audits usually produce noise. **Do Core Web Vitals affect rankings?** Marginally, and they affect conversion considerably. Fix them for the second reason. --- # Share of search and market share URL: https://colleviate.com/blog/share-of-search-vs-market-share.html Topic: Share of search | Published: 2026-10-04 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Share of search tends to lead market share, because search happens during consideration and consideration comes before purchase. The lag is roughly the length of the buying cycle: short in impulse categories, long in considered ones. The relationship breaks down in some categories, so treat the gap between the two numbers as a question to investigate. The claim that search share predicts market share is roughly true, widely overstated, and useful mainly in the gap between the two. The interesting finding behind share of search is not that the two numbers correlate. It is that one of them moves first. In several categories studied by effectiveness researchers, share of search leads market share by months, sometimes by a year or more in considered purchases. ## Why it leads Search happens during consideration, and consideration happens before purchase. In a category where people buy a car every seven years or choose a bank once a decade, the interest is visible long before the transaction. The lag is roughly the length of the buying cycle, which is why it is short in impulse categories and long in considered ones. ## Where the relationship breaks down - Distribution-constrained categories. If the product is not on the shelf, interest does not convert. Share of search will run ahead and stay there. - Price-driven switching. A promotion can take market share without moving brand interest at all. - Brands with ambiguous names, where the search data includes people looking for something else entirely. - Categories bought through intermediaries, where the broker, agent or marketplace is searched instead of the brand. None of these makes the measure useless. They make it a question rather than an answer. ## The gap is the useful part When share of search sits above market share, you have interest you are not converting. That is a distribution, pricing, availability or conversion problem, and it is good news, because those are fixable with things you control. When market share sits above share of search, you are selling more than your brand strength supports. That usually means distribution, promotion or incumbency is carrying you, and it tends to be the position brands are in shortly before a well-funded entrant arrives. It is the single most useful early warning the measure produces. ## How to use it in practice Track both on the same chart, indexed, over three years. Argue about the gap rather than the levels. And resist the temptation to forecast revenue from it: the relationship is directional and category-specific, not a formula, and presenting it as a prediction is how a useful indicator loses its credibility in one board meeting. For what it means for budget, the honest reading is that share of search responds to demand creation rather than demand capture, which is the argument in brand versus performance (https://colleviate.com/blog/brand-versus-performance-split.html). If you want to move it, more retargeting will not do it. ## Questions **Does share of search predict market share?** It leads it in many categories, by roughly the length of the buying cycle. It is directional rather than a formula, and it breaks down where distribution, price or intermediaries dominate. **What does it mean if our search share is higher than our market share?** You have interest you are not converting. Look at availability, pricing and the conversion path before spending more on awareness. **How far ahead does it lead?** Weeks in impulse categories, several months to over a year in considered purchases such as cars, property and financial products. --- # Schema markup that earns results URL: https://colleviate.com/blog/schema-markup-that-earns-results.html Topic: AI visibility | Published: 2026-10-04 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Organization and LocalBusiness on the site, Article or BlogPosting on content, FAQPage where real questions are visible on the page, Product and Service where you sell, and Breadcrumb everywhere. Mark up only what a visitor can see, and never invent ratings or prices. Structured data used to be about getting stars in a search result. It is now about being legible to systems that answer instead of listing. Schema tells a machine what a page is, in a format it does not have to infer. That mattered moderately when the output was ten blue links. It matters more when the output is a generated answer assembled from sources, because the sources that are easiest to parse correctly are the ones most likely to be used. ## What to implement, in order - Organization once, on every page, with the real name, logo, contact and profiles. This is how your brand becomes an entity rather than a string. - LocalBusiness with address, hours and coordinates for anyone with a physical location. - BreadcrumbList on every page below the home page. Cheap, and it changes how the URL displays. - Article or BlogPosting on content, with dates that are true and kept current. - FAQPage where the questions are genuinely on the page and visible to a reader. - Product or Service where you sell, with only the attributes you can evidence. ## The rule that keeps you safe Mark up what is visible. Structured data describing things a visitor cannot see is a policy violation and a manual action risk, and the most common version is invented review counts and ratings. The rich result is tempting. It is not worth the penalty, and reinstatement takes months. ## Where it helps with assistants Clear entity markup, consistent naming across sources, and FAQ content that answers a question in a sentence rather than four paragraphs. The discipline overlaps almost entirely with writing clearly, which we set out in AEO and GEO (https://colleviate.com/blog/aeo-geo-ai-visibility.html). ## Checking it Validate with the rich results test, then watch Search Console for the enhancement reports. A valid markup that produces no enhancement report is usually a type Google does not surface for your page type, which is fine: legibility is still the point. ## What has changed for FAQ and How-to markup? Google limited FAQ rich results to government and health sites in August 2023 and stopped showing them in Search from 7 May 2026; How-to rich results were dropped in 2023. So FAQ and How-to markup no longer earns a search feature, whatever older advice says. The reasoning and an audit are in FAQ schema after the rich result ended (https://colleviate.com/blog/faq-schema-after-the-rich-result-ended.html), and the markup that still describes things Google documents, starting with identity, is covered in Organization schema and sameAs (https://colleviate.com/blog/organization-schema-and-sameas-for-agencies.html). ## Sources - FAQ structured data, documentation and changelog (deprecation notice dated 8 May 2026), Google Search Central, 2026: https://developers.google.com/search/docs/appearance/structured-data/faqpage - Changes to HowTo and FAQ rich results, Google Search Central blog, August 2023: https://developers.google.com/search/blog/2023/08/howto-faq-changes ## Questions **Does schema markup improve rankings?** Not directly. It improves how a page is presented and understood, which affects click-through and how often a page is used as a source. **Can schema markup get a site penalised?** Yes, when it describes content a visitor cannot see. Invented ratings are the most common cause of a structured data manual action. --- # Reading Search Console properly URL: https://colleviate.com/blog/search-console-diagnostics.html Topic: Method | Published: 2026-10-03 | Updated: 2026-11-03 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Impressions move before clicks, so read them first when judging new content. In the indexing report, "crawled, not indexed" and "discovered, not indexed" point at content quality. "Alternative page with proper canonical tag" and "excluded by noindex" are usually intentional and need no action. It is the only free source of truth about how search sees your site, and most teams look at exactly one number in it. Search Console answers three questions: can Google reach the pages, what is it showing them for, and what does it think of them. Most reporting uses only the second, and only the clicks column. ## Impressions before clicks A new page earns impressions before it earns clicks. Filter to the page, look at impressions by query over the first eight weeks, and you will know whether it is being considered for the terms you wanted long before traffic shows up. Judging new content on clicks at week two tells you nothing. ## Indexing statuses worth acting on - Crawled, currently not indexed. Google saw it and declined. Usually thin, duplicative, or not clearly better than what ranks. - Discovered, currently not indexed. Crawl budget or perceived low value. Improve internal linking (https://colleviate.com/blog/internal-linking-for-topic-clusters.html) to it first. - Soft 404. The page looks empty or error-like to a crawler even though it returns a 200. - Redirect error or server error. Fix immediately; these block indexing outright. ## Statuses that are usually fine "Alternative page with proper canonical tag" means a duplicate URL exists and the canonical was honoured, which is the system working. "Excluded by noindex" means you asked for it. "Page with redirect" is a redirect doing its job. These generate notification emails and almost never need action, though redirecting the duplicate is tidier than leaving it. ## Queries you did not target The most useful screen in the tool. Sort by impressions, filter out brand, and look for queries where you appear on page two with a decent impression count. Those are pages that nearly rank, which is the cheapest content work available. ## What it will not tell you Why a competitor outranks you, whether a page is worth having, or how much a ranking is worth. For the first of those, the search results themselves are better evidence than any tool. For the last, see how to measure share of search (https://colleviate.com/blog/how-to-measure-share-of-search.html). Note. The FAQ and How-to reports in Search Console were retired with the features; FAQ schema after the rich result ended (https://colleviate.com/blog/faq-schema-after-the-rich-result-ended.html) explains what changed. Related. A worked example with real numbers: our own Search Console data in the first ninety days (https://colleviate.com/blog/our-own-search-console-first-ninety-days.html). ## Questions **What does "crawled, currently not indexed" mean?** Google fetched the page and chose not to index it. It almost always points at content quality or duplication rather than a technical fault. **Should I worry about "alternative page with proper canonical tag"?** No. It means a duplicate URL was found and your canonical was respected. Redirecting the duplicate is tidier, but nothing is being lost. --- # Branded and category search URL: https://colleviate.com/blog/branded-versus-category-search.html Topic: Share of search | Published: 2026-10-02 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Branded search measures whether people want you; category search measures whether people want what you sell. They move for different reasons and respond to different spending, so adding them together is the most common reporting error in search. Read them separately, give every branded spike a cause, and test brand bidding with a holdout. One tells you whether people want you. The other tells you whether people want what you sell. Teams that merge them end up answering neither question. Branded search is demand that already has your name on it. Category search is demand nobody owns yet. They move for different reasons, they respond to different spending, and the most common reporting error in search is adding them together. ## What branded search tells you That your demand creation is working, or that something happened. It responds to brand campaigns, PR, word of mouth, a product launch and occasionally a crisis, so a spike always needs a cause attached before it is celebrated. It is also the cheapest traffic you will ever get, which creates a familiar argument about whether to bid on your own name. The honest test is a holdout: pause brand bidding for a fortnight and see how much organic absorbs. In most accounts it is most of it. ## What category search tells you The size and direction of the market you are selling into. Rising category search with flat branded search means the category is growing without you, which is the most expensive thing to discover late. It is also where the winnable commercial terms sit. Not the head term, which is enormous and mostly research traffic, but the specific phrases underneath it: the service plus the city, the problem plus the industry, the comparison, the price question. That is the argument in SEO services, what to expect (https://colleviate.com/blog/seo-services-what-to-expect.html). ## Reading them together - Branded up, category flat: you are taking share. The best position to be in. - Category up, branded flat: the market is growing and somebody else is capturing it. - Both down: a category problem, not a marketing one. Check for a substitute. - Branded up sharply with no campaign: find out why before reporting it. It is as likely to be a complaint going around as a win. ## What each should change Branded search informs brand investment and tells you whether the last campaign left anything behind. Category search informs content, SEO and the question of whether to create demand or capture it. Report them as separate lines, always, with the competitive set alongside. A single organic traffic number that mixes both will rise when your brand is in the news and fall when it is not, and nobody will be able to say what to do about it. The full method is in how to measure share of search (https://colleviate.com/blog/how-to-measure-share-of-search.html). ## Questions **Should we bid on our own brand name?** Test it rather than assume. Pause brand bidding for two weeks and measure how much organic absorbs. In most accounts the majority of that traffic arrives anyway. **What does rising category search with flat branded search mean?** The market is growing and competitors are capturing the growth. It is an argument for demand creation rather than more capture spend. **Should branded and non-branded traffic be reported together?** No. They answer different questions and merging them produces a line nobody can act on. --- # Ecommerce SEO for Indian D2C brands URL: https://colleviate.com/blog/ecommerce-seo-india.html Topic: Growth audits | Published: 2026-10-02 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Category pages carry most of the commercial search value in ecommerce, not product pages or the blog. Give each one real content, keep faceted URLs out of the index unless they target genuine demand, and never delete a product page that ranks when the item goes out of stock. The brands that win organic in ecommerce treat category pages as their most important asset. Most treat them as a grid of products with a sentence at the bottom. Search intent in ecommerce is mostly category-shaped. People search "cotton kurta set" far more than they search any individual product, and the page that answers that is your category page. ## Category pages deserve real work A heading that matches the query, a genuine description of what the range is and who it is for, filters that make sense, and internal links (https://colleviate.com/blog/internal-linking-for-topic-clusters.html) to the subcategories underneath. Not two hundred words of keyword text hidden below the fold, which is both obvious and ignored. ## Faceted navigation, controlled Every filter combination can generate a URL, and left alone that produces thousands of near-identical pages competing with each other. Index the facets that match real demand, such as colour or size where people search that way, and keep the rest out with canonicals and parameter handling. ## Product pages Original descriptions, not the manufacturer text every other seller is using. Real measurements. Images that answer the questions that cause returns, which in apparel is most of them, as covered in fashion and retail marketing (https://colleviate.com/blog/fashion-retail-marketing-india.html). Reviews, which add content and the trust signals buyers look for. ## Out of stock is not a reason to delete A ranking product page that disappears takes its position with it. Keep the page, say it is unavailable, show alternatives, and offer a notification. Deleting it and redirecting to the category is the second-best option; a 404 is the worst. ## Competing with marketplaces Marketplaces will often outrank you for your own product names, and you will not beat them on domain strength. You can beat them on the things they cannot carry: full range pages, buying guides, sizing and care content, and the brand story. That is also the argument for building the direct channel at all, set out in marketplace or your own store (https://colleviate.com/blog/marketplace-or-own-store.html). ## Questions **What should an ecommerce category page include?** A heading matching the search, a real description of the range, sensible filters, and internal links to subcategories. It carries more commercial search value than any product page. **What should we do with out-of-stock product pages?** Keep them live, mark them unavailable, show alternatives and offer notification. Deleting a ranking page gives away the position. --- # Fintech marketing inside the rules URL: https://colleviate.com/blog/fintech-marketing-india.html Topic: Growth audits | Published: 2026-10-01 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Fintech marketing in India works inside the rules by treating regulation as a design constraint. Returns cannot be promised, risk disclosures must be clear, and the regulated entity behind the product must be named. Because users are asked to share money and identity with a new brand, trust is the main conversion lever, and content does unusual work. Financial marketing is the only category where the regulator reads your creative. Treat that as a design constraint and the work gets better, not worse. Most fintech marketing problems look like creative problems and are actually trust problems. Somebody is being asked to connect a bank account, share identity documents or borrow money from a brand they met ninety seconds ago. Everything in the funnel is downstream of whether that feels safe. ## The constraints are real, and they are useful Returns cannot be promised. Risk disclosures are not optional and cannot be hidden in grey four point type. Lending communication carries rules on how rates and charges are expressed, and the regulated entity behind the product has to be named, which matters because a great many consumer apps are the front end for somebody else's licence. Agencies new to the category treat this as a tax on the idea. The better reading is that it removes the lazy options. You cannot promise outcomes, so you have to be interesting about the mechanism, the design, the service, or the honesty itself. ## Trust is the conversion lever - Name the regulated entity and the licence plainly. Burying it signals exactly what you hoped it would hide. - Show the charges before the signup, including the ones that appear later. Fee surprises are the single largest driver of early churn and of the complaints that end up on review sites. - Put a human somewhere. A named support route, a real address, a response time you honour. - Treat security copy as product copy. How data is held and what happens if something goes wrong. These are not brand garnish. In most fintech funnels we have looked at, moving the fee disclosure earlier lowers signups and raises activated users, which is the trade you want. The landing page mechanics are in landing pages that convert (https://colleviate.com/blog/landing-pages-that-convert.html). ## The acquisition maths Fintech makes it unusually easy to buy growth that is not growth. Installs, signups and even completed KYC can all rise while the number that matters stays flat. Define the activated user precisely, which is usually the first real transaction or the first funded account, and report cost per activated user rather than cost per lead. Most paid channels look twice as good on the earlier metric and half as good on the right one. Then look at retention by cohort before scaling anything. A channel that acquires users who transact once is a subsidy programme with a dashboard. ## Content does more work here than almost anywhere People research money decisions. Clear explanatory content ranks, gets cited, and shortens the trust gap before the first ad is ever seen. It also feeds the AI answers that increasingly sit between a question and a product. The discipline is the same as in content that earns links (https://colleviate.com/blog/content-that-earns-links.html), with one addition: every factual claim needs a source and a review date, because a stale rate in an explainer is a compliance exposure as well as an error. ## Plan for the bad week Outages happen, charges get disputed, and a single viral complaint can undo a quarter of acquisition. Have the holding statement, the escalation route and the named spokesperson agreed before you need them. The work is cheap in advance and impossible to do well under pressure: see crisis communications (https://colleviate.com/blog/crisis-communications.html). ## Questions **What can fintech brands not say in India?** Guaranteed or assured returns, misleading comparisons, and anything that obscures charges or risk. Lending communication must express rates and fees clearly, and the regulated entity behind the product has to be identified. **What is the right acquisition metric for fintech?** Cost per activated user, where activation is a first real transaction or a funded account. Cost per install and cost per signup both flatter channels that deliver nobody who transacts. **Does content marketing work for financial products?** It is one of the strongest channels in the category, because people research money decisions. Every claim needs a source and a review date, since stale figures are a compliance exposure. --- # Taking an Indian brand into overseas search URL: https://colleviate.com/blog/international-seo-for-indian-brands.html Topic: Method | Published: 2026-10-01 | Updated: 2026-10-01 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Use subdirectories on one strong domain for most cases, add hreflang only when you have genuinely different pages per market, and localise currency, shipping, contact and proof before translating anything. One market done properly beats five done thinly. The technical part is well documented and rarely the thing that fails. What fails is launching five markets at once with the same page and a different flag. Overseas search is a distribution question before it is a technical one. The question is not how to implement hreflang, it is whether the market has demand you can serve profitably and what a buyer there needs to see before trusting an unfamiliar brand. ## Domain structure Subdirectories on one domain concentrate authority and are the right default for most brands. Country domains are stronger locally and far more expensive to build, since each one starts from nothing. Subdomains sit awkwardly between the two. Pick subdirectories unless you have a specific legal or operational reason not to. ## Hreflang, only when it is earned Hreflang tells search engines which version of a page belongs to which language and region. It is useful when you genuinely have different pages per market. It is pointless, and a source of errors, when the only difference is a currency symbol. Implement it reciprocally or not at all, because one-directional tags are ignored. ## Localise the things that matter first - Currency, pricing and whether tax is included. - Shipping times, duties and returns, which are the questions that actually stop a first purchase. - A contact route in that market, even if it is just an email answered in local hours. - Proof the buyer recognises: local reviews, local press, local payment methods. ## Content is not translation A translated page ranks for translated phrases, which may not be what people in that market search. Do keyword research in the target language before translating, and expect the best-performing page in one market to be the wrong page in another. ## The mistake that costs six months Launching several markets simultaneously with thin localisation. Each one performs badly, none produces enough data to learn from, and the conclusion drawn is that international does not work. One market, properly served, will teach you what the next needs. ## Questions **Subdirectory or country domain for international SEO?** Subdirectories on one strong domain in most cases. Country domains perform better locally but start from zero authority and cost far more to maintain. **Do we need hreflang?** Only when you have genuinely different pages per market. If the only difference is currency, it adds errors without adding value. --- # Healthcare marketing without overclaiming URL: https://colleviate.com/blog/healthcare-marketing-india.html Topic: Growth audits | Published: 2026-09-30 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Healthcare marketing in India has to avoid overclaiming: law restricts cure claims, professional councils limit how practitioners advertise, and superlatives, cure claims and patient testimonials create most problems. The job is to be the clearest, most verifiable option for a worried person comparing a few providers, using search, careful review handling and privacy-respecting measurement. Health marketing is governed by advertising law, professional codes and the fact that the audience is frightened. Only the first two are usually planned for. Somebody searching for a hospital at eleven at night is not a funnel. They are worried, they are comparing three options, and they will make a decision on evidence they can verify quickly. The marketing job is to be the clearest, most verifiable option on the screen. ## The rules, briefly Indian law restricts claims about curing a long list of conditions, and professional councils limit how practitioners may advertise, including the use of testimonials and before and after imagery in several contexts. Drug and device promotion carries its own constraints. The practical effect is that superlatives, cure claims and patient testimonials are the three things most likely to create an issue, and they are the three things a generalist agency reaches for first. What remains is more durable anyway: credentials, outcomes you can evidence, facilities, accreditation, and the specific experience of being a patient there. ## What patients actually search Not your hospital name. Symptoms, procedures, costs, recovery times, insurance coverage and whether a particular doctor is available. That is a content map, and most health websites answer almost none of it. A procedure page that states what it involves, who it is for, what it costs in a range, how long recovery takes and what the risks are will outrank and outconvert a page describing how patient centric you are. Local search does the rest of the work, because healthcare decisions are geographic. Profile completeness, accurate hours, department listings and a steady flow of recent reviews decide visibility: see local SEO for service businesses (https://colleviate.com/blog/local-seo-for-service-businesses.html). ## Reviews, carefully You cannot write them, you cannot filter them by expected sentiment, and in health the replies are read very closely. Reply without confirming or denying anyone's treatment, because acknowledging that a named person was a patient is a confidentiality problem even when the review says so. Move specifics to a private channel and say so publicly. The reputation mechanics are in reputation is a search result (https://colleviate.com/blog/reputation-is-a-search-result.html). ## Measurement that respects privacy Health interest is sensitive data. Do not pass condition names, procedure pages or appointment details into advertising platforms through URLs, event names or custom audiences. Separate the analytics layer from the advertising layer deliberately, review what your tags actually transmit, and keep the consent position defensible under the data protection rules. The implementation discipline is in a GA4 setup that survives scrutiny (https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html). ## What works - Doctor-led explanatory content, in plain language, by someone with a name and a qualification. - Price and process transparency, which almost nobody offers and which patients consistently say they want. - Accreditation and outcome data, stated without decoration. - Speed of response. An appointment request answered in ten minutes beats a brand campaign. None of that requires a claim anyone could challenge, which is the point. In a category where the regulator and the patient are both reading, the compliant version of the work is usually also the persuasive one. ## Questions **Can hospitals advertise in India?** Yes, within limits. Cure claims for restricted conditions are prohibited, professional councils constrain how practitioners advertise, and testimonials carry risk. Credentials, facilities, accreditation and process transparency are safe and effective. **Can we use patient testimonials?** Treat them as high risk. Several professional codes restrict them, and even a public reply confirming someone was a patient raises a confidentiality problem. Outcome data and accreditation do the same job more safely. **How should healthcare sites handle tracking?** Keep condition and procedure signals out of advertising platforms entirely. Separate analytics from advertising tags, audit what is transmitted, and keep consent defensible. --- # Share of search in B2B URL: https://colleviate.com/blog/share-of-search-for-b2b.html Topic: Share of search | Published: 2026-09-30 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Share of search still works in B2B, with adjustments for thin data: use absolute volumes from a keyword tool, aggregate monthly or quarterly, widen the term set with product names and misspellings, and cross-check with Search Console. The searcher is often not the decision-maker, so search leads by the whole buying process. If volumes stay too small, consideration is happening elsewhere. The measure was built on consumer categories with plenty of volume. B2B has neither, and it is still worth doing, with adjustments. Most B2B brands discover the same thing when they first try this: the volumes are too low for Google Trends to return a usable series. That is a data problem, not a reason to abandon the question. ## Working around thin data - Use absolute volumes from a keyword tool rather than relative interest. Small numbers are still comparable to each other. - Aggregate monthly or quarterly. Weekly data in a low-volume category is almost entirely noise. - Widen the term set to include product names and common misspellings, which in technical categories carry real volume. - Use Search Console impressions for your own brand as a cross-check. It is first-party and it does not round to zero. If after all that the numbers are still too small, the honest conclusion is that search is not where your category’s consideration happens yet, and the budget argument should move to where it does. ## The buying group problem In B2B, the person searching is often not the person deciding. An analyst researches, a manager shortlists, a committee approves and a procurement team negotiates. Search volume reflects the first of those, which means it leads by the length of the whole process, frequently a year or more. That makes it a genuinely long leading indicator, and it also makes it a poor basis for quarterly decisions. Read it annually, alongside pipeline source data: see B2B lead generation that sales will accept (https://colleviate.com/blog/b2b-lead-generation-india.html). ## Category creation changes the question If you are selling something the market has no name for, category search barely exists and share of search measures nothing useful. In that situation, track the problem language instead: the symptoms people search when they do not yet know your category exists. Growth in that language is the leading indicator, and your branded share of it is the real measure of whether the category story is landing. ## What it will not tell you Deal quality. Account penetration. Whether the right people inside a target account have heard of you, which in enterprise is the question that actually matters and which search cannot answer. Pair it with something that can: an annual awareness survey of named accounts is cheap, and it measures the buying group (https://colleviate.com/blog/how-b2b-buyers-shortlist-agencies.html) rather than the researcher. ## Questions **Our B2B search volumes are too low for Google Trends. What now?** Use absolute volumes from a keyword tool, aggregate quarterly, widen the term set, and cross-check with Search Console impressions for your own brand. **How far ahead does share of search lead in B2B?** Often a year or more, because the person searching sits at the start of a long buying process involving several other people. **Does it work for a new category?** Not directly. Track the problem language people search before they know your category exists, and your share of that instead. --- # Short-form video that holds attention URL: https://colleviate.com/blog/short-form-video-that-works.html Topic: Creative | Published: 2026-09-30 | Updated: 2026-09-30 Publisher: Colleviate (https://colleviate.com/) **Short answer.** The first frame decides everything. Lead with the most interesting moment rather than building to it, caption everything because most people watch muted, keep one idea per video, and cut the moment the idea lands rather than filling a target length. Everybody knows the first three seconds matter. Almost nobody edits as though they believe it. Short video is not a shorter film. It is a different form with its own grammar, and the most common failure is treating it as an ad with the beginning trimmed off. ## Open at the interesting part Whatever moment you were building towards, start there. The setup can follow once somebody has decided to stay. A brand logo, a slow pan or a title card in the first second is a decision to lose most of the audience before the idea arrives. ## One idea per video Three points in forty seconds means none of them lands. Make three videos. They cost almost nothing extra if you shot sensibly, and each one gets its own chance in the feed. ## Caption everything Most feed viewing is muted. Burned-in captions, legible at phone size, in a face that survives compression. Auto-captions need correcting, especially with Indian names, places and any Hinglish, where the errors are both frequent and funny in ways nobody intended. ## Length is an outcome, not a target Cut when the idea is finished. A nine-second video that completes beats a thirty-second video people abandon, because completion and rewatch drive distribution far more than duration does. ## Making enough of it The constraint for most teams is supply, not ideas. Shoot in blocks: one session producing a month of material, with a consistent setup so the edits are fast. Film the long-form thing you were making anyway and cut it down, which is the argument for filming the podcast in are branded podcasts worth it (https://colleviate.com/blog/branded-podcasts-worth-it.html). ## What not to copy Trending audio with no connection to what you sell. A format that worked for a creator whose whole appeal is their face, when you have no face. And anything that requires the viewer to already know your brand, which in a feed is almost nobody. ## Questions **How long should a short-form video be?** As long as the idea needs and no longer. Completion rate drives distribution, so a nine-second video that finishes beats a thirty-second one people leave. **Do we need captions on social video?** Yes. Most feed viewing is muted, and auto-captions need correcting, particularly for Indian names and Hinglish. --- # How to write a marketing plan for the year URL: https://colleviate.com/blog/annual-marketing-plan.html Topic: Method | Published: 2026-09-29 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A useful annual marketing plan is a short set of decisions, not a list of activity with a budget attached. Start with the diagnosis: not enough people know you, they know you and choose someone else, or they choose you and leave. Then set targets you can defend, review quarterly, and write it clearly enough that a sceptical finance director can disagree with it. Most annual plans are a list of activity with a budget attached. The useful ones are a short set of decisions, written down clearly enough that somebody can disagree with them. A plan earns its place by making the next twelve months easier to argue about. If it cannot be contradicted, it is a wish list. The test is simple: can a sceptical finance director read it and point to the thing they think is wrong? ## Start with the diagnosis, not the calendar Before any channel appears, state what the problem actually is. Not enough people know us. People know us and choose someone else. People choose us and leave. These are three different businesses and they need three different plans. Writing the diagnosis down first prevents the most common failure, which is a plan full of tactics that address whichever problem is most fun to solve. The evidence for the diagnosis should be in the document. Search trends, win and loss reasons, repeat rates, what customers say when asked to describe you without prompting. An outside-in audit (https://colleviate.com/blog/outside-in-brand-audit.html) is the fastest way to assemble it if nobody has done so recently. ## The four decisions - Who we are for, and who we are not for. The second half is the one that changes anything. - What we want people to believe that they do not currently believe. - Where the money goes, split between creating demand and capturing it. - What we will stop doing. A plan with no subtractions is a budget request. ## Targets you can defend Set one business target, two or three marketing targets that plausibly cause it, and a leading indicator for each. Pipeline or revenue is the business target. Qualified enquiries, cost per acquisition and repeat rate are marketing targets. Share of search, branded query volume and content citations are leading indicators, which matter because they move months before the others do. Resist the urge to set a target for everything. A plan with fourteen KPIs has none, because nobody can tell which one you would protect in a bad quarter. ## Quarterly, not annual Write the year, commit the quarter. The first quarter should be specific down to the work. The remaining three should state intent and budget, with the detail filled in as evidence arrives. Teams that fully specify twelve months spend the second half defending decisions made with worse information than they now have. Review quarterly against the leading indicators. Change the tactics freely. Change the diagnosis only when something real has changed, because a diagnosis rewritten every quarter is not a diagnosis, it is a mood. ## What to put in the document Keep it short enough to be read. One page of diagnosis with the evidence. One page of decisions. One page of the quarter ahead. One page of budget and what each tranche buys. A measurement appendix for anyone who wants it. If the plan runs to forty slides, the thinking is somewhere inside it, and nobody will find it. ## The part that gets skipped Write down what would have to be true for you to change course. Name the number, the date and the decision. This is the single most useful paragraph in any plan, because it converts the inevitable mid-year argument from a debate about confidence into a check against something you already agreed. It also protects good work from being cancelled in month four, which is the fate of most brand investment: see brand versus performance (https://colleviate.com/blog/brand-versus-performance-split.html). ## Questions **How long should a marketing plan be?** Short enough that the people approving it read all of it. Four pages plus a measurement appendix is usually plenty; a forty slide deck hides the thinking rather than showing it. **How detailed should the later quarters be?** Intent and budget only. Fully specifying twelve months commits you to decisions made with the worst information you will ever have on them. **What is the most common mistake?** Starting from channels. A plan that opens with a channel list has skipped the diagnosis, and will spend the year solving whichever problem was easiest to describe. --- # Community management when the comments are hostile URL: https://colleviate.com/blog/community-management-india.html Topic: Creative | Published: 2026-09-29 | Updated: 2026-09-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Reply to genuine questions and genuine complaints, publicly and quickly. Do not reply to bait. Take specifics to DM and say publicly that you have. In a pile-on, one clear statement beats fifty individual replies, and silence after that is a decision you review hourly rather than a default. The replies are read by far more people than the person you are replying to. That one fact should decide how you write them. Community management gets treated as an intern job until the week it becomes the most consequential thing the brand is doing. It is customer service performed in front of an audience, and the audience is the point. ## What to answer Real questions, always, and fast. Real complaints, publicly, with an apology if one is owed and a route to resolution. Compliments, briefly, because it signals a human is present. Not bait. Not an argument somebody is clearly enjoying. Not a competitor. Engaging with those feeds the thing you want to shrink, and the audience can tell the difference between a brand being helpful and a brand being defensive. ## Speed matters more than polish An acknowledgement in twenty minutes beats a perfect reply at six the next evening. In India, where a large share of this arrives through Instagram DMs and WhatsApp rather than comments, response time is close to the whole of the service experience. ## Move specifics, say you moved them Order numbers, refunds, anything personal goes to DM. Say so in the public thread, so everyone reading knows it was handled rather than ignored. Never ask for private details in a public comment. ## When it becomes a pile-on Stop replying individually. Fifty variations of the same apology reads as a bot and extends the story. Post one clear statement, pin it, and route everything to it. If the underlying complaint is real, fix it before saying it is fixed, which is the discipline in what to do in the first six hours (https://colleviate.com/blog/crisis-communications.html). ## The thing nobody schedules time for Reading the comments for what they actually say. Patterns in complaints are the cheapest product research available, and in most companies they stop at the social team because nobody has asked for them. ## Questions **Should brands reply to every comment?** Reply to genuine questions and complaints. Replying to bait extends the thread and reads as defensive to everyone else watching. **How fast should a brand reply on social?** Within the hour during working hours. In India, where much of this arrives by DM, response time is most of the service experience. --- # Fashion and retail marketing that moves stock URL: https://colleviate.com/blog/fashion-retail-marketing-india.html Topic: Creative | Published: 2026-09-28 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Fashion marketing that moves stock escapes the discount cycle by changing what the discount is for: clear old season deliberately, in a separate place, on a known schedule, and keep new season at full price without exception. Drops work because they are scarce, imagery is the product page, and returns are a marketing problem worth measuring. Fashion is the category where marketing and inventory are the same conversation, and where the easiest growth lever quietly destroys the business. Apparel has a structural trap. Discounting works immediately, it is always available, and every use of it teaches customers to wait. A brand two years into that habit has trained its entire base to buy only at sale, which shows up as revenue growth and margin collapse at the same time. ## Escaping the discount cycle Not by refusing to discount, which ignores the reality of seasonal stock. By changing what the discount is for. Clear old season deliberately and visibly, in a separate place, on a schedule people understand. Keep new season at full price without exception. The brands that hold this line for two years recover their full price share; the ones that run a sitewide offer every six weeks never do. The number to watch is the proportion of revenue taken at full price. It is more honest than revenue growth and it is the first thing that moves when the brand is weakening, as in luxury brand marketing (https://colleviate.com/blog/luxury-brand-marketing-india.html). ## Drops work because they are scarce, not because they are frequent A drop model gives you a reason to communicate, a reason to buy now, and a rhythm the audience can learn. It stops working the moment everything is a drop. Fewer, better, on a predictable cadence, with genuinely limited quantity, beats a weekly launch of things that stay available. ## Imagery is the product page Nobody can touch the fabric. The photography has to do the work of the fitting room, which means it has to answer the questions that cause returns: how it actually sits, on more than one body, in daylight, with the texture visible and the length stated. Styling imagery sells the idea. Detail imagery sells the size, and the second kind is the one most brands skip. Video does more here than almost any category, because movement answers the drape question that stills cannot. Shoot both in the same session rather than commissioning twice: see what video production costs in India (https://colleviate.com/blog/video-production-cost-india.html). ## Returns are a marketing problem In Indian apparel, return rates routinely reach a third of orders and higher on marketplaces, and the cost lands on contribution rather than on revenue. Most of it traces to sizing uncertainty and imagery that oversells. A size guide with real measurements, consistent fit across the range, the length and model height stated on every product, and honest colour reproduction will cut returns more than any campaign will raise orders. ## What to measure Full price share, sell through by collection, return rate by product and by channel, repeat purchase rate within twelve months, and contribution after returns and acquisition cost. Revenue alone hides every one of the problems above, and in fashion all of them compound quietly before they appear. ## Questions **How do fashion brands reduce returns?** Real measurements, consistent fit across the range, model height and garment length on every product, honest colour, and detail imagery. Sizing uncertainty causes most returns, and photography that oversells causes the rest. **Is discounting always bad in fashion?** No, but it should be structural: old season, in a separate place, on a predictable schedule. Sitewide offers every few weeks train the base to buy only at sale. **What is the most useful metric in apparel?** Full price share of revenue, read alongside sell through and return rate. It moves before the brand problem shows up in the top line. --- # Share of search in the Indian market URL: https://colleviate.com/blog/share-of-search-in-india.html Topic: Share of search | Published: 2026-09-28 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Share of search transfers to India, but the term list must not be English-only. Include Hinglish and transliterated spellings, taken from your own Search Console data, and regional-language queries. Separate festive seasonality from real trend, and remember that discovery often starts on marketplaces, YouTube and quick commerce apps, so search engine volume understates total demand. The method transfers. The term list does not, and a brand measuring Indian demand with an English-only keyword set is reading a fraction of its market. Four things make Indian search demand different enough to change how the measure is built. None of them is exotic, and all of them are routinely missed by tracking set up from a global template. ## Transliteration and Hinglish A large share of queries are Hindi or another Indian language typed in Roman script. People search “sasta” alongside “cheap”, “ghar” alongside “home”, and brand names get spelled several ways. If your term list contains only the English spelling of your brand and category, you are counting a subset and comparing it against competitors whose names happen to transliterate more consistently. The fix is unglamorous: collect the variants from your own Search Console data, where the real spellings appear, and build the set from what people actually type. ## Regional language volume is growing faster than English Devanagari and other script queries are a growing share of total volume, and in several categories they now dominate outside the metros. A measure that excludes them will show a brand looking healthy in English while losing the larger market. Track at least the two or three languages that matter for your geography, as separate lines rather than merged. ## Festive seasonality is not a trend Indian category demand is shaped by a festive calendar that moves year to year, so a simple year-on-year comparison will mislead whenever the dates shift. Compare like periods relative to the festival rather than to the calendar date, and smooth over a full year before declaring anything about the trend. The same discipline applies to wedding seasons, exam cycles and the monsoon in the categories they govern. ## Discovery does not start on Google alone For a growing share of buyers the first search happens on a marketplace, on YouTube, on a quick commerce app or inside a messaging thread. Search engine volume still works as a comparative measure across brands, because the bias applies to everyone in the set, but it understates total demand. Where marketplace search share is available, it is worth tracking alongside, especially in D2C. ## What to do with it Build the term set from first-party data, include transliterations and the main regional languages, compare festive periods to festive periods, and read the series annually. Then use it for the thing it is good at: telling you whether interest in your brand is growing faster or slower than interest in the people you compete with, which is the argument in share of search as a leading indicator (https://colleviate.com/blog/share-of-search-leading-indicator.html). ## How large and how multilingual is India's search audience? The IAMAI and Kantar Internet in India report put active internet users at 886 million in 2024, with rural users, at about 488 million, outnumbering urban users. The same report found that 98 percent of internet users accessed content in Indic languages. That is the scale argument for everything above: a share of search built on English spellings alone is measuring a minority of the market's searching. ## What does it show in Indian categories? We built it for nine categories in the Colleviate Share of Search Index (https://colleviate.com/blog/share-of-search-index-india.html), with the method and data published. Three readings show why the measure earns its place in India. Groww has led Zerodha in stockbroking search every month since September 2025. Birla Opus went from no measurable search interest to 11.0 percent of decorative paints, ahead of two long-established brands. And in mobile networks, BSNL briefly led prepaid recharge searches in July 2024, the month the private operators raised tariffs, which is a reminder that a spike needs a cause before it is read as a trend. ## How do you build an Indian term list? Start from your own data, not a keyword tool. Export the queries from Search Console that include your brand and your competitors' names and read them: the transliterated spellings, the Hinglish phrasings and the misspellings your buyers actually type are all there. Group the variants of each name into one term, test each term's related queries in Google Trends to make sure it means what you think, and add regional-language forms for the states that matter to you. Then decide what search engine data cannot see. If your buyers start on a marketplace or a quick-commerce app, track your position there separately and read both together. Our India index (https://colleviate.com/blog/share-of-search-index-india.html) publishes its term list and every exclusion, which is the standard worth holding your own tracking to. ## Sources - India’s internet revolution: key insights from Kantar and IAMAI report, Business Today, 16 January 2025: https://www.businesstoday.in/amp/technology/news/story/indias-internet-revolution-key-insights-from-kantar-and-iamai-report-461043-2025-01-16 - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **Should Indian brands track Hinglish search terms?** Yes. A large share of queries are Indian languages typed in Roman script. Build the term list from your own Search Console data rather than from assumptions. **How do you handle festive seasonality?** Compare festival period to festival period rather than calendar date to calendar date, since the dates move, and smooth across a full year before reading a trend. **Does search share still work when discovery happens on marketplaces?** As a comparative measure, yes, because the same bias applies to every brand in the set. It understates total demand, so track marketplace search alongside it where you can. --- # What a brand owes a festival URL: https://colleviate.com/blog/what-a-brand-owes-a-festival.html Topic: Culture | Published: 2026-09-28 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A brand has standing in a festival when it contributes something to it: a product people actually use then, a genuine role in how it is celebrated, or money and effort put somewhere useful. Without one of those, a greeting post is wallpaper and an emotional film is borrowing. Every October the feed fills with diyas, and almost none of it is about the brand posting it. Indian festive marketing runs on an assumption nobody examines: that a brand is entitled to participate in a cultural moment simply because its audience is observing it. Sometimes true. Mostly not. ## Standing has to come from somewhere Three things give a brand the right to speak in a festival. It makes something people genuinely use then. It plays a real part in how the occasion works, like travel, gifting, food or clothing. Or it does something materially useful, which means spending money rather than running a hashtag. A software company posting a rangoli has none of those. It is not offensive, it is simply noise, and it costs the brand a small amount of the credibility it spent the year accumulating. ## Why the emotional film keeps failing The template is familiar: a family separated, a journey, a reunion, a slow track, a logo. It is made because it is safe, and it is forgotten because forty other brands made it the same week. Nothing in it could only have come from this brand. The ones that last do something specific: a real observation about how the festival is actually experienced now, by the people living it rather than the people remembering it. That is harder and it is the entire job. ## India is not one festive calendar Onam, Durga Puja, Pongal, Baisakhi, Eid, Christmas in Goa and Kerala, and a dozen regional observances that matter more locally than the national ones a Mumbai agency defaults to. A brand with real regional presence and no regional recognition of it is telling its customers something about how closely it is paying attention. ## The cost of getting it wrong Appropriation reads instantly to the people whose festival it is, and it travels fast. The test is simple: would somebody who grew up with this recognise themselves in it, or recognise a stock image of themselves? Our longer argument is in what a campaign owes the culture it borrows from (https://colleviate.com/blog/what-a-campaign-owes-the-culture.html). Related. The practical plan: how to plan a festive campaign (https://colleviate.com/blog/how-to-plan-a-festive-campaign.html). Related. How advertising has borrowed from the national calendar before: iconic Indian ads for the ages (https://colleviate.com/blog/iconic-indian-ads-for-the-ages.html). ## Questions **Should every brand run a festive campaign?** No. Without a product people use then, a genuine role in the occasion, or something materially useful to contribute, a festive post is wallpaper. **Why do festive brand films all feel similar?** Because they are built from the same safe template. The ones that are remembered contain a specific observation that could only have come from that brand. --- # Why your ads stopped working URL: https://colleviate.com/blog/why-your-ads-stopped-working.html Topic: Measurement | Published: 2026-09-27 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Ads usually stop working for one of five reasons, each with its own fingerprint: creative fatigue, audience exhaustion, broken tracking, competitive pressure or a decayed offer. Creative fatigue, for example, shows as rising frequency and falling click-through rate with stable cost per click. Diagnose which one before changing anything. Performance rarely collapses for the reason the account manager gives first. Five things cause almost all of it, and they leave different fingerprints. The conversation usually starts the same way. The account was fine in March, it is bad now, nothing changed. Something changed. The useful work is finding out which of five things it was, because the fixes have nothing in common. ## 1. Creative fatigue The fingerprint: frequency climbing, click-through rate falling, cost per click stable. The same people have seen the same asset too many times. The fix is new creative, which sounds obvious and is almost never done at the required rate. On an active social account, the main assets need refreshing every four to six weeks, and that means several distinct ideas rather than recolours of one. Most brands produce too little creative and then spend the shortfall on audience experiments. ## 2. Audience exhaustion The fingerprint: reach plateauing, cost per acquisition rising steadily rather than suddenly, and the best performing audiences getting smaller. You have harvested the cheap demand. This is not a problem with the account, it is the account working exactly as designed until the available intent ran out. The fix is upstream: create demand rather than capture more of it. If this is the diagnosis, no amount of optimisation will reverse it, which is the argument in brand versus performance (https://colleviate.com/blog/brand-versus-performance-split.html). ## 3. Tracking broke The fingerprint: a step change on a specific date, often after a site release, a consent banner change or a tag edit. Conversions reported fall while the finance system shows no such fall. Check this first, always, because it is free to check and it invalidates every other diagnosis. A platform optimising towards a broken signal degrades fast and silently. See a GA4 setup that survives scrutiny (https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html). ## 4. Competitive pressure The fingerprint: cost per click rising across the board, impression share falling, no change in your own metrics otherwise. Somebody funded entered your auction. Nothing is wrong with your account. The options are narrow: pay more, target more precisely, or compete somewhere they are not. Of those, the third is usually the one nobody considers. ## 5. The offer decayed The fingerprint: clicks holding, conversion rate falling, sales reporting more price objections. This is the uncomfortable one, because it is not a marketing problem. A competitor improved, a price moved, a review cycle turned. Paid media is the first place it shows up and the last place it can be fixed. The landing page is worth checking first, since it is cheap to improve and frequently the actual constraint: see landing pages that convert (https://colleviate.com/blog/landing-pages-that-convert.html). ## How to run the diagnosis Pull twelve months, not four weeks. Look at frequency, impression share, click-through rate, conversion rate and cost per click as separate lines, because the aggregate number hides which one moved. Mark every tracking or site change on the chart. Then compare against the finance system, since a decline that exists only in the platform is a measurement problem wearing a performance problem's clothes. Most accounts we look at have two of these at once. The order of repair matters: fix tracking, then creative, then the offer, then the structural question of where demand comes from. ## Questions **How often should ad creative be refreshed?** On an active social account, every four to six weeks, with several distinct ideas rather than variations of one. Search creative lasts longer; the landing page is usually the constraint there. **Our costs rose but sales did not fall. What happened?** Usually competitive pressure or a tracking change. Compare platform-reported conversions against the finance system before changing anything in the account. **Can a new agency fix rising acquisition costs?** Only if the cause is creative, structure or tracking. If the cause is exhausted demand or a weakened offer, a new agency will reach the same ceiling a quarter later. --- # Marketing in a language you do not speak URL: https://colleviate.com/blog/regional-language-marketing-india.html Topic: Culture | Published: 2026-09-27 | Updated: 2026-09-27 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Translate the intent, not the words. Brief a writer who works in that language rather than a translator working from English copy, cast voices from the region rather than approximating the accent, and put approval in the hands of someone who speaks it, not someone reading a back-translation. The English line is written, approved, and then sent for translation. Everything that goes wrong afterwards was decided in that sentence. Most multilingual campaigns in India are English campaigns wearing other languages. The idea was built in English, the rhythm is English, the pun only works in English, and the regional versions arrive as obligations rather than work anybody wanted to make. ## Translation and transcreation are not the same job Translation carries the words. Transcreation carries the intent, and usually discards the words to do it. A line that rhymes in Hindi will not rhyme in Tamil, and forcing it produces copy that native speakers find stilted in a way an English-speaking approver cannot detect. Brief the idea, not the sentence. Give a regional writer the proposition and the thing it has to make somebody feel, and let them write it. ## Dialect is a real decision Hindi in Delhi is not Hindi in Bhopal or Patna. Tamil for Chennai is not Tamil for Madurai. Pick deliberately based on where the business is, and know that choosing the broadcast-standard version is itself a choice that sounds a particular way to a local ear. ## Cast from the region An approximated accent is the fastest way to tell an audience you were not really speaking to them. This is also where the money is well spent: the right voice carries a mediocre script, and the wrong voice sinks a good one. ## Approval, honestly Back-translation into English tells you the meaning survived. It tells you nothing about whether the line is any good, or whether it is accidentally comic. Build a review step with someone who speaks the language and has permission to say the work is weak, not just that it is accurate. ## What to test Comprehension, tone and whether anything lands wrong, with ten people from the actual audience rather than ten colleagues. For search, the term list is a separate problem, covered in share of search in the Indian market (https://colleviate.com/blog/share-of-search-in-india.html). ## Questions **Is translation enough for regional campaigns?** Rarely. Translation carries words; transcreation carries intent and usually rewrites the line. Brief a regional writer with the idea rather than the English sentence. **How do we approve work in a language nobody senior speaks?** Add a reviewer who speaks it and has authority to call the work weak, not just accurate. Back-translation checks meaning, not quality. --- # Education marketing for a long decision URL: https://colleviate.com/blog/education-marketing-india.html Topic: Growth audits | Published: 2026-09-26 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Education is sold to the student and bought by the parent, so campaigns need two jobs: aspiration for the student, evidence and reduced risk for the parent. Measure enquiry quality rather than volume, treat the counselling conversation as the funnel, and plan around a decision cycle long enough that most campaigns never see their own result. Education is sold to one person and bought by another, over a cycle long enough that most marketing teams never see the end of their own campaign. The student wants a future they can picture. The parent wants risk removed. Those are different briefs, and a campaign that speaks only to one of them produces enquiries the other will veto, usually after your sales team has spent three weeks on them. ## Two audiences, two jobs To the student: what you will become, who you will be among, what the place feels like. Aspiration carried by people who are visibly like them, not by a building shot at golden hour. To the parent: outcomes, placements, safety, faculty credentials, fees stated plainly, and the sense that this is a decision nobody will regret. Evidence rather than atmosphere. Run both deliberately. The most common failure is a campaign built entirely on aspiration, with the evidence buried three clicks deep on a page the parent never reaches. ## Enquiry quality, not enquiry volume Admissions teams drown in enquiries that were never serious. The fix is the same as in other high-consideration categories: qualify in the advertising rather than after it. Show the fee range, the eligibility, the location and the intake. Enquiries will fall and applications will not, which is the trade that matters. The same logic applies in real estate marketing (https://colleviate.com/blog/real-estate-marketing-india.html), for the same reason. ## The counselling conversation is the funnel In education, the decisive moment is a human one: a call, a campus visit, an open day. Marketing exists to produce those and to prepare people for them. That makes response time and counsellor quality more consequential than any creative decision, and it makes the handover between marketing and admissions the place where most budget leaks. Measure it directly. Enquiry to contact, contact to counselling session, session to application, application to enrolment. Most institutions can name only the first and the last. ## Content is the long game and the cheap one Course comparison pages, fee and scholarship explainers, placement data, curriculum detail, visa and hostel information. These rank for exactly what prospective students search months before they enquire, they get shared between parent and child, and they go on working for years. Nearly every institution underinvests here while buying the same lead ads as everyone else. The approach is in content that earns links (https://colleviate.com/blog/content-that-earns-links.html). ## The cycle governs everything Admissions is seasonal and the season is fixed. Brand work happens in the off cycle, because it needs time to accumulate. Capture spends hard in the intake window, when intent is concentrated. Planning this a year ahead is the difference between buying attention at a sensible price and bidding against every competitor in the same six weeks: see how to write a marketing plan for the year (https://colleviate.com/blog/annual-marketing-plan.html). ## Questions **Should education marketing target students or parents?** Both, with different material. Aspiration for the student, evidence of outcomes and safety for the parent. Campaigns built only on aspiration produce enquiries the parent later vetoes. **How do we improve admissions enquiry quality?** Qualify in the advertising: fees, eligibility, location and intake. Volume falls, applications hold, and the counselling team stops spending its week on people who were never eligible. **What should education marketing measure?** Every step of the funnel: enquiry to contact, contact to counselling, counselling to application, application to enrolment. Most institutions track only the first and the last. --- # How to brief a photoshoot URL: https://colleviate.com/blog/photography-direction-brief.html Topic: Creative | Published: 2026-09-26 | Updated: 2026-09-26 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A usable brief states what each image is for, where it will appear and at what crop, the shot list with counts, the styling and casting decisions already made, and the usage rights required. References should show the specific quality you want, with a note saying which quality. Reshoots are almost never caused by the photographer. They are caused by a brief that left the decision to the day. A shoot is a few hours where everything costs money simultaneously. Every decision deferred to that morning is made under time pressure by whoever is standing closest. ## Say what each image is for Hero banner, product listing, social square, print, email header. Each has a crop, a safe area and a resolution. A beautiful horizontal frame is useless to a team that needed vertical, and this gets discovered a week later. ## References, specified Collecting twenty pretty images is not a reference pack. For each one say what you are pointing at: this lighting, this crop, this expression, this background, this colour. Unlabelled references create confident misinterpretation, because everybody sees a different thing in the same picture. ## The shot list, with counts How many setups, how many looks, how many products, how many variations of each. Then check it against the day. Fifteen setups and four looks is not one day, and the difference between the list and the clock is where the quality goes. ## Decide styling and casting beforehand Props, surfaces, wardrobe, hands, skin tones, hair. Approve them before the day, with substitutions agreed. On casting, pick for fit with the actual audience rather than for conventional appeal, which we argue in casting for evidence (https://colleviate.com/blog/casting-for-evidence.html). ## Usage, in writing Where the images will run, in which territories, for how long. Talent fees and licensing depend on it, and extending afterwards always costs a multiple of buying it at the start. The same economics are in what video production costs in India (https://colleviate.com/blog/video-production-cost-india.html). ## The question that prevents most reshoots Ask the photographer what they would need to know to shoot it without you present. The gaps in their answer are the gaps in your brief. ## Questions **What should a photography brief include?** Purpose and crop for each image, a shot list with counts, styling and casting decided in advance, references with notes on what each one demonstrates, and the usage rights required. **How many setups fit in a shoot day?** Fewer than most briefs assume. Count the setups against the hours, including lighting changes and styling resets, before committing to a single day. --- # PR worth paying for URL: https://colleviate.com/blog/pr-worth-paying-for.html Topic: Culture | Published: 2026-09-25 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** PR is worth paying for when it delivers credibility, discovery and corroboration, not coverage volume. Corroboration matters more than ever, because search results, procurement checks and AI answers treat a claim as settled when independent editorial sources repeat it. Judge a retainer on a few meaningful placements and refuse paid placements sold as coverage. The PR industry is measured by the wrong number almost everywhere, and the wrong number is coverage volume. A monthly report listing forty placements looks like value. Look closely and thirty of them are the same press release reprinted by sites nobody reads, three are paid placements labelled as coverage, and one is the piece that actually did something. ## What PR is for Three outcomes justify the spend. Credibility, which is somebody with editorial standards vouching for you in front of their audience. Discovery, which is reaching people who would never see your ads. And corroboration, which is the quiet one: the reason a search result, a procurement check or an AI answer treats your claim as settled rather than as marketing. That third outcome has become more valuable than the industry has noticed. Assistants and search engines weight consistency across independent sources, so a handful of credible references change what gets said about you in places you cannot edit. We set out the mechanism in AEO and GEO (https://colleviate.com/blog/aeo-geo-ai-visibility.html). ## What a good retainer looks like - A point of view to sell, not a stream of announcements. Journalists do not need your news; they need a story your news belongs in. - Named target publications agreed at the start, with a realistic view of which are reachable this year. - Access to someone worth quoting. PR fails most often because the people with opinions are unavailable for comment. - Preparedness work, which is the unglamorous half: the position on the awkward question, agreed before it is asked. ## What to refuse Guaranteed placements, which are advertising with the label removed. Reporting by advertising value equivalent, a metric the industry itself has disavowed and which still appears in decks. Volume targets, which incentivise exactly the wrong behaviour. And any retainer where nobody can tell you which journalist covered your category last month. ## Measuring it honestly Count placements in publications you named, not in aggregate. Track referring domains earned, since a credible article links. Watch branded search in the week after a placement, which is the cheapest read on whether anyone noticed. Ask sales whether a piece shortened a conversation. And over a longer horizon, check what the first page of results says about you, because that is what a buyer sees before the first call. What not to do is attribute revenue to coverage directly. PR is an upstream investment with an uncertain lag, and forcing it into a last click model produces a number that is both precise and false. ## The uncomfortable part Most companies are not interesting enough to earn coverage on request. That is not a failure of the agency. The remedy is to make something worth reporting: original data, a position somebody will argue with, a decision other companies are facing. An agency that tells you this on day one is more useful than one that promises forty placements a month, and considerably rarer. ## Questions **How much does a PR retainer cost in India?** Boutique retainers commonly run from about Rs 1 lakh to Rs 5 lakh a month depending on scope and seniority. Volume-led agencies quote less and deliver reprints. **How long before PR shows results?** Two to three months before the first meaningful placements, longer for national business press. Relationships and a credible point of view take a quarter to build. **Is paid placement worth it?** Sometimes, as advertising. It should never be reported as earned coverage, and it carries little of the corroboration value that makes PR worth the money. --- # How to name a brand URL: https://colleviate.com/blog/brand-naming.html Topic: Creative | Published: 2026-09-25 | Updated: 2026-09-25 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Generate far more candidates than feels sensible, screen early for trademark class conflicts and domain availability because that removes most of them, then test the survivors for pronunciation, spelling from hearing it, and unfortunate meanings in the languages of your market. Naming feels like the creative part and is mostly an elimination process. The good ones survive the screening, they are rarely the ones that win the room. A name does less work than people expect and lasts longer than anything else you will decide. It does not have to explain the business. It has to be sayable, spellable, available and capable of carrying meaning once you have put meaning into it. ## Generate wastefully A hundred candidates is a normal starting point. Real words, invented words, compounds, words from other languages, founders' references, places. The first twenty will be descriptive and most will be unusable. ## Screen before you fall in love Trademark search in the relevant classes, in every market you intend to operate in. Domain availability, including what you would pay for the one you want. Social handles. A plain search for existing companies and, importantly, for anything embarrassing the name already means somewhere. This stage kills most candidates, which is why generating wastefully matters. Doing it after the team has chosen a favourite is how brands end up with a name plus a vowel removed. ## Descriptive names are a trap They explain the business on day one and constrain it on day one thousand. They are also hard to protect legally and impossible to own in search, because the words belong to the category. A slightly abstract name takes longer to mean something and can eventually mean exactly what you build. ## Test the practical things Say it on a phone call. Spell it to someone who has only heard it. Check how it reads in the languages of your market, where an innocent English word can land badly. Write it small next to the logo. These are not glamorous tests and they are the ones that surface the real problems. ## Then stop No name sounds right at first. Recognition is built, not found, and a committee searching for the name that already feels familiar will choose the most generic candidate every time. Pick the one that survives the screening and can be held, then hold it, which is the argument in audacity and tenacity (https://colleviate.com/blog/audacity-and-tenacity.html). ## Questions **Should a brand name describe what the company does?** Usually not. Descriptive names constrain the business later, are hard to protect legally and are impossible to own in search. **When should trademark checks happen in naming?** Early, before anybody chooses a favourite. Trademark and domain screening removes most candidates, which is why you generate far more than you need. --- # How to launch a brand properly URL: https://colleviate.com/blog/brand-launch-campaign-plan.html Topic: Creative | Published: 2026-09-24 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A brand launch is the first three months of a position you intend to hold, not an announcement. Settle what the brand stands for before launch day, do the quiet work of press, spokespeople and seeding in the four weeks before, concentrate paid reach in the first fortnight, then sustain. Judge it on whether anything changed after the noise. A launch is not an announcement. It is the first three months of a position you intend to hold, and the announcement is the least important part of it. Most launches are built backwards. The date is set, the film is made, the influencers are booked, and the question of what the brand actually stands for is settled somewhere in the edit. Then the noise happens, and two months later nothing has changed except the budget. ## What has to be true before launch day - The position is written down in a sentence somebody could disagree with. If it could belong to a competitor, it is not a position. - The distinctive assets exist. Not a logo. A colour, a shape, a phrase, a sound, a layout, applied consistently from day one, because recognition starts accumulating at the first exposure. - The destination works. The site, the product page, the enquiry route and the tracking, tested on a mid range phone. Driving launch traffic to an unfinished destination wastes the one moment when attention is free. - The proof is ready. Whatever substantiates the claim: the data, the demonstration, the first customers, the certification. ## Sequencing the ninety days Weeks minus four to zero. Quiet work. Brief the press who actually cover your category, prepare the people who will speak, seed to a small group who can say something credible, and get the destination finished. Earned coverage needs lead time and almost never arrives on request. Weeks one to two. The loud part. Paid reach, coverage, creators, whatever the plan contains, concentrated rather than spread. A launch that trickles out over six weeks reads as a series of unconnected posts. Weeks three to twelve. The part everyone skips. Keep the same idea in market, in different executions, with the budget tapering rather than stopping. This is where the launch either becomes a brand or becomes a memory. The argument is in audacity and tenacity (https://colleviate.com/blog/audacity-and-tenacity.html). ## Where the money goes Roughly: a third on making the thing properly, because a weak asset cannot be rescued by spend. A third on the concentrated first fortnight. A third held for weeks three to twelve, protected in writing so it does not get spent early when the launch week numbers look exciting. Resist spending the entire budget on day one. The most common launch failure is a brilliant week followed by silence, and silence is where the recognition you just bought goes to die. ## What tells you it worked Not launch day impressions. Branded search volume four weeks later, compared with the baseline you recorded beforehand. Direct traffic. Unprompted mentions. Whether anyone can describe the proposition back to you without prompting. Revenue matters, obviously, and in most categories it arrives after the signals above have already moved. Record the baseline before you launch. Almost nobody does, and without it every subsequent claim about the launch is unfalsifiable. ## The mistake worth naming Launching the brand and the product at the same time as the website rebuild and the new positioning, all in one week, so that when something works nobody can tell which thing it was. Stagger them. The learning is worth more than the tidiness. ## Questions **How long should a brand launch campaign run?** Plan for ninety days, not a week. Concentrate the first fortnight, then keep the same idea in market with a tapering budget through week twelve. **How much of the budget should go to the launch week?** About a third, with a third on production and a third protected for the two months after. Spending everything on week one buys a spike and loses the recognition it created. **What should we measure after a launch?** Branded search, direct traffic and unprompted mentions against a baseline recorded before launch. Without that baseline, nothing you claim afterwards can be checked. --- # A tone of voice people can actually use URL: https://colleviate.com/blog/tone-of-voice-that-survives.html Topic: Creative | Published: 2026-09-24 | Updated: 2026-09-24 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Replace adjectives with rules. "Confident, friendly, human" changes nothing. "We never use exclamation marks", "we say price not investment", and a rewritten before-and-after for each rule will change what people write tomorrow. Most voice documents are three adjectives and a mood board, which is why most brands sound like each other. The test of a voice guideline is whether a freelancer who has never met you can write a passable paragraph from it. Most fail, because they describe a personality rather than giving instructions. ## Adjectives are not rules Confident, approachable, bold, human. Every brand claims these and none of them tells a writer what to do on a Tuesday. A rule is specific enough to be broken: we do not use exclamation marks, we write "you" not "customers", we never start a sentence with "At Colleviate, we believe". ## Show the rewrite For every rule, a before and after using your own real copy. This is the part people actually read and the part most documents omit. Twenty pairs teaches a voice faster than ten pages of description. ## Decide the boring things Numerals or words. Oxford comma or not. How prices are written. Whether you capitalise product names. Date format. British or American spelling, which in India is worth stating because both arrive daily. These are not style preferences, they are the difference between a site that reads as one organisation and one that reads as eleven freelancers. ## Write the "never" list The words you will not use. Journey, solutions, leverage, unlock, seamless, passionate, ecosystem, game-changing. A short prohibition list does more for a brand's writing than any amount of aspirational description. ## Keeping it alive A voice survives through editing, not documents. Someone has to read what goes out and send things back. Where that person does not exist, the voice reverts to the industry default within about two quarters, regardless of how good the guideline is. ## Questions **What makes a tone of voice guideline usable?** Specific rules with before-and-after rewrites from your own copy, decisions on the mechanical questions, and a list of words you will not use. **Who keeps a brand voice consistent?** An editor. Documents describe a voice; editing maintains one. Without somebody sending work back, a voice drifts to the category default within a couple of quarters. --- # Are branded podcasts worth it? URL: https://colleviate.com/blog/branded-podcasts-worth-it.html Topic: Creative | Published: 2026-09-23 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A branded podcast is worth it in a narrow set of situations: a long sales cycle where trust matters more than reach, access to guests the audience wants to hear, and a subject deep enough for fifty episodes. You want at least two of the three. Podcasts are cheap to start and expensive to sustain, so plan for the long run. Podcasts are cheap to start, expensive to sustain, and valuable in a narrow set of situations that are easy to identify in advance. The audio business runs on a misleading average. A handful of shows have enormous audiences, most have a few hundred listeners, and the brands launching into it usually model the first group while landing in the second. ## When it works Three conditions, and you want at least two of them. A long sales cycle where trust matters more than reach, so a thousand committed listeners are worth more than a million impressions. Access to people your audience would want to hear from, which is a real asset and often the only genuine differentiator. And a subject with enough depth to sustain fifty episodes, because the format rewards accumulation and punishes novelty. Professional services, B2B software, finance, healthcare and categories where the buyer is researching for months all fit. Impulse consumer categories almost never do. ## When it does not If the honest reason is that a competitor launched one, stop. If nobody internally wants to host it, stop, because an uninterested host is audible within thirty seconds. And if the plan depends on reaching a large audience quickly, stop, since organic audio discovery is slow and the shows that grew fast almost all imported an audience from somewhere else. ## The real cost Recording is the cheap part. The cost is in the production discipline: booking guests, preparing properly, editing to a standard people will tolerate, writing the show notes, cutting the clips, and publishing on the same day every time for a year. Budget for an editor and a producer, not for a microphone. Video changes the maths. Filming the conversation gives you the episode plus a month of short-form, which is usually where most of the audience actually meets the show. If you are producing anyway, film it: the incremental cost is small against what the clips return. On how production economics work more generally, see what video production costs in India (https://colleviate.com/blog/video-production-cost-india.html). ## Formats that survive - The access interview. People your audience cannot easily get to, asked questions they would ask. - The teardown. A recurring analysis of something real, which gives the show a reason to exist beyond the guest list. - The internal expert series. Cheaper, more sustainable, and better suited to categories where the knowledge is genuinely scarce. What dies quickly is the general business conversation with no particular angle, which is most of them. ## Measuring it Downloads are a vanity number past the first month. Track completion rate, which tells you whether the format works. Track returning listeners, which tells you whether the show works. Track whether guests share the episode, which is free distribution and a good proxy for quality. And track the thing that usually justifies the whole exercise, which is whether the sales team can point to a deal where a prospect mentioned it. In most B2B cases, the business case rests on a small number of listeners who matter rather than a large number who do not. That is a perfectly good case. It is just not the one the download chart tells. ## Questions **How many episodes before a branded podcast works?** Plan for at least twenty before judging it, published on a reliable schedule. Shows abandoned at episode eight were never given the conditions to work. **Should we film the podcast?** If the budget allows it, yes. The short-form clips usually reach more people than the audio episodes, and the incremental production cost is small. **What should a branded podcast cost?** Sustainable production with editing, clips and show notes commonly runs from about Rs 50,000 to Rs 2 lakh an episode depending on format and whether it is filmed. --- # Packaging is the only medium they hold URL: https://colleviate.com/blog/packaging-as-media.html Topic: Creative | Published: 2026-09-23 | Updated: 2026-09-23 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Packaging has to work at three distances: recognisable across a shop floor or a thumbnail, legible at arm reach, and informative in the hand. Design for the worst reproduction it will face, which for most brands is a compressed marketplace thumbnail. Every other channel competes for attention. Packaging already has it, in both hands, at the moment of use. It is also the only brand asset a customer pays for, keeps for a while and sometimes photographs. Treating it as a production detail rather than a channel is a strange decision that most brands make by default. ## Three distances At five metres, or as a thumbnail, only colour and shape register. At arm reach, the name and the one thing that distinguishes it. In the hand, the detail: what it is, what is in it, how to use it. Designs that look exquisite flat on a screen routinely fail the first distance, which is the one that decides whether anybody picks it up. ## Design for the worst reproduction A marketplace thumbnail at low resolution, a phone photograph under shop lighting, a courier box that arrived scuffed. Fine rules, thin type and subtle tonal shifts disappear in all three. The print conditions matter too: colours shift between substrates, and a brand colour approved on coated stock will not match on kraft. ## The unboxing question Elaborate unboxing is a real driver of sharing in some categories and pure cost in most. The test is whether anyone would film it unprompted. If not, spend the money on better materials and less of them, which also avoids the growing customer impatience with packaging waste. ## Constraints to design within - What the filling line can actually run, which is not everything a designer can draw. - Regulatory text, which in food, cosmetics and pharma is non-negotiable and larger than anyone plans for. - Shipping: what survives a courier, and whether the box ships flat. - Cost per unit at real volume, not at sample volume. ## Where it pays back Distinctiveness, mostly. A pack recognisable at thumbnail size is doing media work every time it appears in a feed, a marketplace grid or somebody else's photograph, and it keeps doing it long after a campaign ends. ## Questions **How much does packaging affect sales?** Heavily at the moment of choice, because it is the last communication before the decision. It also functions as media every time the product appears in a photograph or a marketplace grid. **Is elaborate unboxing worth the cost?** Only where customers would film it unprompted. Otherwise spend on better materials and less of them, which also reduces waste complaints. --- # What a website should cost URL: https://colleviate.com/blog/website-cost-india.html Topic: Method | Published: 2026-09-22 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A website in India typically costs about Rs 1.5 lakh to Rs 5 lakh for a credible small business site on a good template with original content, and about Rs 6 lakh to Rs 20 lakh for a custom designed brand site. Ecommerce or application-like sites should be scoped as software. Budget for ongoing costs and own your domain, code and content. Website quotes in India range from forty thousand rupees to forty lakh for what sounds like the same brief. Both ends can be honest, and the brief is why. A website is not one thing. It is design, content, build, integrations, infrastructure and the ongoing cost of keeping it alive. Quotes differ because suppliers assume different answers to questions the brief left open. ## What drives the number - How much is designed rather than chosen. A template configured well is a legitimate product. A designed site costs several times more because somebody draws every state of every component. - Page count and template count. Twenty pages built from four templates is cheap. Twenty bespoke layouts is not. - Who writes the words. Content is the most underestimated line in every website project, and the most common reason launches slip by months. - Integrations. Payments, CRM, inventory, login, multi language, anything connecting to a system somebody else controls. - Photography and film. Usually a separate budget, frequently discovered late. ## Rough bands A credible small business site on a good template, with original content and sensible technical foundations, commonly runs from about Rs 1.5 lakh to Rs 5 lakh. A designed site for a brand that cares what it looks like, with custom components and real content work, from Rs 6 lakh to Rs 20 lakh. Ecommerce with integrations, or a site that is effectively an application, goes well beyond that and should be scoped as software rather than as design. Below about a lakh you are buying a template someone else filled in. That is sometimes exactly the right purchase, and it should be described honestly rather than sold as design. ## The costs that arrive later Hosting and domain. Licences for the theme, plugins, fonts and stock. Maintenance and security updates, which are not optional on a content system. Changes after launch, which always come, and which cost a fortune when there is no retainer and no handover documentation. Ask for the annual running cost in writing before signing anything; a cheap build with an expensive tail is the standard shape of a bad deal. ## What you must own The domain, registered to you. The hosting account. The analytics property. The design source files. The content. The code, or at least a copy of it. Agencies that hold these do not usually intend to hold you hostage, and it works out that way regardless when you leave. ## Comparing quotes honestly Ask every supplier for the same four things: the number of unique templates, who writes the content, what the annual running cost will be, and what happens to the site if you stop working together. The quotes become comparable immediately, and the gap between them usually turns out to be scope rather than price. One more thing worth paying for, which almost no quote includes by default: the migration plan. If the site is replacing an existing one, the redirects and the preserved content decide whether you keep your search traffic, as set out in redesigning without losing your rankings (https://colleviate.com/blog/website-redesign-without-losing-rankings.html). ## Questions **How much does a business website cost in India?** A credible template-based site with original content commonly runs Rs 1.5 to 5 lakh. A designed site for a brand that cares how it looks runs Rs 6 to 20 lakh. Ecommerce and application-like sites cost more and should be scoped as software. **Is a template site bad?** No. A well configured template with original content and sound technical foundations beats a weak custom build. The problem is paying custom prices for it. **What should we own at the end of a website project?** Domain, hosting account, analytics property, design source files, content and the code. Get this in writing before the project starts. --- # Marketing an Indian SaaS product URL: https://colleviate.com/blog/saas-marketing-india.html Topic: Growth audits | Published: 2026-09-22 | Updated: 2026-09-22 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Decide early whether you are selling to India or from India, because the two need different pricing, proof and channels. Early pipeline comes from founder-led outbound, communities and documentation-grade content, not from paid media, which works later once the message is proven. The hardest question is not which channel. It is whether the buyer is in Bengaluru or Boston, because almost everything downstream depends on the answer. Indian software companies sit in an unusual position: a domestic market that is large, price-sensitive and relationship-driven, and an export market that pays more and demands more proof. Trying to serve both with one set of materials produces work that convinces neither. ## Selling to India Longer relationship cycles, more negotiation, real sensitivity to annual commitments, and a strong preference for seeing the product work on the customer's own data before committing. Pricing pages tend to underperform against a conversation. Referral and reputation inside an industry do most of the work. ## Selling from India Credibility is the constraint. An overseas buyer evaluating an unfamiliar vendor looks for security documentation, support hours in their timezone, named customers they recognise, and signs of permanence. Those are marketing assets as much as operational ones, and they belong on the site rather than in a sales deck. ## Trial or demo A free trial works when the product delivers visible value alone within a session. A demo works when configuration or data is required before anything is useful. Offering both without a clear route usually means prospects pick the one that lets them avoid talking to you, then churn out of the trial unhelped. ## What actually produces early pipeline - Founder-led outbound to a list of companies you can name and explain. - Documentation and technical content so good that users send it to each other. - Communities where your buyers already argue about the problem. - Comparison and alternative pages, which capture buyers already in the market. Paid media rarely starts the engine. It scales one that is already turning, which is the argument in B2B lead generation that sales will accept (https://colleviate.com/blog/b2b-lead-generation-india.html). ## The measurement problem Long cycles mean quarterly reporting is mostly noise. Track pipeline created and influenced, cycle length by source, and the leading indicators of category interest, which for a new category means problem-language search rather than product search: see share of search in B2B (https://colleviate.com/blog/share-of-search-for-b2b.html). ## Questions **Should an Indian SaaS company target India or export markets?** Choose deliberately and early. Domestic selling is relationship-led and price-sensitive; export selling is credibility-led. One set of materials rarely serves both. **Does paid media work for early-stage SaaS?** Rarely as a starting engine. It scales a motion that already converts; before that it buys traffic against a message that has not been proven. --- # Marketplace or your own store? URL: https://colleviate.com/blog/marketplace-or-own-store.html Topic: Growth audits | Published: 2026-09-21 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Marketplace or own store is really a question of who owns the customer. A marketplace gives you demand you did not create at an all-in cost well above the headline commission; your own store gives you a relationship you can use again, but you pay for traffic. Most brands need both, and should compare contribution per order after acquisition. The argument is usually about margin. It is really about who owns the customer, and what that ownership is worth to you in three years. A marketplace gives you demand you did not create. Your own store gives you a relationship you can use again. Most brands need both, and the mistake is not choosing one, it is drifting into dependence on whichever one grew fastest first. ## What each channel actually costs On a marketplace, add commission, fulfilment, storage, returns, advertising inside the platform, and the discounts the platform expects during sale events. The all-in figure routinely lands far above the headline commission, and most of it is not negotiable. On your own store, the fees are lower and the demand is not free. You pay for traffic, and you pay for the technology and the people who run it. The honest comparison is contribution per order after acquisition cost, not margin before it. ## The ownership question This is where the two channels really differ. On a marketplace you get an order; the platform gets a customer. You cannot email them, you usually cannot see who they are, and you are competing on a page that also shows four alternatives and a cheaper private label version of your product. On your own store you get both. That is only worth something if you then do something with it, which most brands do not. If your lifecycle programme consists of two discount emails a quarter, the ownership advantage is theoretical. The mechanics are in email and lifecycle marketing (https://colleviate.com/blog/email-and-lifecycle-marketing.html). ## Where each one wins - Marketplaces win on discovery for products people search by category rather than by brand, on heavy or low-value items where fulfilment matters, and in cities where your own delivery promise is weak. - Your own store wins on margin, on bundles and subscriptions, on anything requiring explanation, and on every repeat purchase you can earn directly. ## Running both without being eaten Keep pricing consistent enough that the marketplace does not train your customers to wait for a sale event. Hold back something for your own store, whether that is a bundle, a size, a colour or a subscription, so there is a reason to buy direct. Treat marketplace advertising as a cost of being in that channel rather than as your growth engine. And watch the ratio: once a single platform passes about half your revenue, your pricing and your roadmap are being set by somebody else. ## Quick commerce changes the question For impulse and replenishment categories, ten minute delivery platforms have become a third channel with their own economics: high velocity, thin margin, brutal listing competition, and almost no customer data. They are worth entering for the velocity and the visibility. They are a poor foundation for a brand, for the same reason the other marketplaces are, only faster. ## The decision Start where the demand already exists, because early revenue buys you time. Build the direct channel in parallel rather than later, because later is always after the dependence is set. And measure the two on contribution after acquisition cost, not on gross margin, which is the number that makes marketplaces look worse and your own store look better than either deserves. The wider economics are in D2C growth after the first crore (https://colleviate.com/blog/d2c-growth-after-the-first-crore.html). ## Questions **What percentage of revenue should come from marketplaces?** There is no fixed answer, but once one platform passes roughly half your revenue, it is setting your pricing and your roadmap. Treat that as a risk threshold rather than a success. **Is it cheaper to sell on our own site?** Lower fees, yes. Cheaper overall only if you can acquire traffic efficiently. Compare contribution after acquisition cost, not margin before it. **Should a new brand launch on a marketplace first?** Often yes, for the demand and the speed. Build the direct channel alongside it rather than afterwards, because dependence is easier to avoid than to unwind. --- # Automotive marketing and the dealership gap URL: https://colleviate.com/blog/automotive-marketing-india.html Topic: Growth audits | Published: 2026-09-21 | Updated: 2026-09-21 Publisher: Colleviate (https://colleviate.com/) **Short answer.** The constraint is almost never lead volume, it is the handover between the digital enquiry and the dealership. Measure cost per test drive and cost per retail rather than cost per lead, and fix response time before increasing media spend. The brand spends nationally, the dealer sells locally, and the enquiry falls into the space between them. Car buying in India is researched online for months and completed in a showroom. That split creates the category's defining problem: the people spending the media budget and the people closing the sale work for different companies with different incentives. ## Where the leads die An enquiry arrives at the dealership, sits in a spreadsheet, and is called the next afternoon by somebody with forty others. The buyer, meanwhile, enquired with three brands in the same hour. Response time decides the outcome more reliably than creative, targeting or offer. This is an operations problem that marketing budget cannot solve, and it is usually cheaper to fix. The same pattern governs property, as set out in real estate marketing that produces site visits (https://colleviate.com/blog/real-estate-marketing-india.html). ## Measure deeper than the lead Cost per lead flatters everything. Price the funnel to cost per contactable lead, cost per test drive and cost per retail. Reallocate on the last two. The sources that look expensive per lead are frequently the cheapest per sale. ## What buyers actually research Mileage and running cost, service network reach, resale value, waiting period, and ownership reviews from people who have had the vehicle a year. Brand films do not answer any of those, and a brand that answers them properly on its own site intercepts the research instead of leaving it to forums and reviewers. ## Electric adds three anxieties Range in real conditions, charging access on the routes the buyer actually drives, and battery life with the cost of replacement. These are the objections, and evasive messaging about sustainability does not touch them. Specific, honest answers do. ## Local is the whole last mile Dealership listings, accurate hours, service reviews and map visibility decide which showroom gets the visit. Most national campaigns underinvest here while competing hard for the national search term: see local SEO for service businesses (https://colleviate.com/blog/local-seo-for-service-businesses.html). ## Questions **What is the most important metric in automotive marketing?** Cost per test drive and cost per retail. Cost per lead hides the handover losses where most of the budget actually leaks. **How quickly should automotive enquiries be contacted?** Within minutes. Buyers enquire with several brands in one session, and the first credible response usually gets the test drive. --- # AI in marketing: what actually works URL: https://colleviate.com/blog/ai-in-marketing-what-works.html Topic: AI visibility | Published: 2026-09-20 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** AI tools genuinely help with the mechanical parts of marketing work, but the gains sit in a narrower band than the sales material suggests. They quietly degrade anything that needs a point of view, because they reproduce the consensus, and they produce confident wrong numbers that must be checked. Judge a tool by which parts of the work get better and which get worse. The useful question is not whether to use these tools. It is which parts of the work get better when a machine does them, and which parts get quietly worse. Two years of enthusiasm have produced a reasonably clear picture. The gains are real and they are concentrated in a narrower band than the sales material suggests. The losses are real too, and they show up later, which is why they are underreported. ## Where it genuinely helps - Volume work with a verifiable output. Resizing and versioning, transcription, first-pass translation, alt text, product descriptions from structured data, cut-downs from a long edit. - Research compression. Reading a hundred reviews, clustering open-ended survey answers, summarising a category, pulling the recurring objection out of sales call transcripts. - Drafting, where someone who knows the subject then rewrites it. The draft is scaffolding, not output. - Analysis you would not otherwise do. Anomaly checks across campaigns, pattern finding in performance data, the second look nobody has time for. ## Where it quietly degrades the work Anything that needs a point of view. These models produce the consensus of what has been written. In a crowded category the consensus is precisely what everybody else is already saying, which is why AI-written thought leadership reads as competent and changes nothing. Facts you do not check. Confident wrong numbers are the characteristic failure, and in regulated categories a wrong number is a compliance event rather than an embarrassment. Volume for its own sake. Publishing forty thin pages a month because it is now cheap produces a site search engines treat as low value and buyers ignore. The economics of content did not change; only the cost of the first draft did. The standard is still the one in content that earns links (https://colleviate.com/blog/content-that-earns-links.html). ## Disclosure and risk Synthetic faces and voices in advertising carry real exposure: likeness rights, platform policy and the reputational cost of being caught presenting a generated person as a customer. Do not present generated imagery as documentary. Do not use a synthetic voice resembling a real person without permission. Keep a record of what was generated and with which tool, because you will eventually be asked. ## The other side of the coin AI is also now a distribution channel rather than only a production tool. A growing share of buyers ask an assistant before they search, and what it says about you is assembled from sources you do not control. That is a marketing problem with a marketing answer, which we set out in AEO and GEO (https://colleviate.com/blog/aeo-geo-ai-visibility.html) and in how answer engines choose sources (https://colleviate.com/blog/how-ai-answer-engines-choose-sources.html). ## How to judge a tool Ask what it replaces and what it adds to the review queue. A tool that produces ten times the output and requires a human to check every item has moved the bottleneck rather than removed it. Ask where the data goes, which matters for client confidentiality. And run it against work you have already done, where you know what good looks like, before you run it against work you cannot yet evaluate. ## Questions **Does AI-written content rank?** It can, and it rarely earns links or citations, because it contains nothing a model could not already produce. Use it for drafting and structure, never for the claims. **Where does AI help most in marketing?** Volume work with verifiable output, research compression, first drafts that an expert rewrites, and analysis nobody currently has time to do. **Do we need to disclose AI-generated imagery?** Do not present generated imagery as documentary, and do not use synthetic likenesses or voices without permission. Keep a record of what was generated and how. --- # Wellness marketing without the claims you cannot make URL: https://colleviate.com/blog/wellness-and-fitness-marketing.html Topic: Growth audits | Published: 2026-09-20 | Updated: 2026-09-20 Publisher: Colleviate (https://colleviate.com/) **Short answer.** The economics are retention, not acquisition: a member who stays eleven months is worth several who join on a discount and leave in six weeks. Claims about results, weight loss or health outcomes are regulated, so build the marketing on the experience and the community instead. The category sells transformation and is not allowed to promise it. That constraint is the whole creative brief. Fitness and wellness businesses run on recurring revenue, which means the marketing problem is almost never getting people through the door. It is what happens in week three. ## Retention is the business A gym with strong acquisition and poor retention is a leaking bucket with an expensive tap. The levers are operational more than promotional: onboarding in the first fortnight, a named person who notices when somebody stops coming, progress people can see, and community that makes absence socially expensive. Marketing supports that with lifecycle communication rather than offers, which is the mechanics in email and lifecycle marketing (https://colleviate.com/blog/email-and-lifecycle-marketing.html). ## Discounting trains the wrong member Deep joining offers attract people buying a price rather than a habit, and they churn fastest. They also devalue the membership for everyone paying full rate, who eventually find out. Run acquisition offers rarely, structurally, and preferably as a trial rather than a discount on the annual. ## Claims, carefully Results claims, before-and-after imagery and anything implying a health outcome are regulated territory in India, and supplements sit under food regulations with specific labelling and claim rules. Build on what you can evidence: the facility, the coaching, the programme design, the people who train there. ## Trainers are the channel In most studios the trainers have more audience and more trust than the brand does. That is an asset if you support it and a risk if you ignore it, because when a trainer leaves, their members often go too. Make them visible, give them content support, and accept that you are partly in the business of marketing individuals. ## Local search decides the trial Gym choice is geographic and fast. Profile completeness, class timings, recent photographs and a steady flow of reviews determine who gets the walk-in, well before any campaign does. ## Questions **Can fitness brands show before-and-after images in India?** Treat them as high risk. Results and health-outcome claims are regulated, and supplements carry additional food-labelling rules. Build on facility, coaching and community instead. **Why do joining discounts hurt gyms?** They attract members buying a price rather than a habit, who churn fastest, and they devalue the membership for everyone paying full rate. --- # Luxury brand marketing in India URL: https://colleviate.com/blog/luxury-brand-marketing-india.html Topic: Creative | Published: 2026-09-19 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Luxury is a belief held partly by people who will never buy, which makes ownership mean something. Every tactic that widens the funnel cheaply, such as discount codes, flash sales and aggressive retargeting, spends a little of that asset. In India many premium brands run a D2C performance playbook and wonder why the brand feels thinner each year. Expensive is a price. Luxury is a belief held by people who are not buying, which is why the discipline is so unforgiving of shortcuts. The Indian premium market has grown faster than the craft of selling into it. Plenty of brands have the price, the packaging and the store. Rather fewer have the thing that makes a price feel inevitable rather than optimistic. ## The uncomfortable rule A luxury brand is partly defined by the people who admire it and will never own it. That is what makes ownership mean something. It also means every tactic that widens the funnel cheaply, the discount code, the flash sale, the aggressive retargeting, spends a little of the asset to buy a little revenue. Most premium brands in India are running a performance playbook borrowed from D2C and wondering why the brand feels thinner each year. ## What actually builds it - Provenance, stated specifically. Who makes it, where, how long it takes, what is difficult about it. Vague heritage copy is worthless; a named workshop and a real process are not. - Restraint in distribution. Where you are available is a brand decision before it is a sales decision. Availability everywhere is the fastest way to stop being special. - Craft in the work itself. Photography, film, print, the weight of the packaging, the person at the counter. In this category the execution is the argument, which is why production quality is not a line item to trim. - Patience. Luxury is built by holding a position for a decade. This is the argument in audacity and tenacity (https://colleviate.com/blog/audacity-and-tenacity.html), and nowhere does it apply more strictly. ## The Indian specifics Three things make this market its own problem. Discounting is culturally normal, including at the top end, so scarcity has to be designed rather than assumed. The buyer is frequently younger than the global average and researches heavily online before entering a store, so the digital surface has to carry the same craft as the boutique. And the gifting and wedding economy distorts seasonality in ways that reward preparation and punish improvisation. ## Where digital fits Search and social matter, but the job is different. The role of digital in luxury is to be found in impeccable condition, not to chase conversions. That means the site performs, the imagery is yours, the story is complete, and the enquiry is answered by a person quickly. Paid media is best used for reach among a defined audience and for retargeting that stops short of pestering. The thing to avoid is the discount ladder: the escalating offer sequence that works in D2C and, in premium, teaches your most valuable customers that the price is negotiable. Once they learn that, you cannot unteach it. ## Measuring it Revenue tells you about the quarter. For the brand, track unprompted recall among people who cannot currently afford you, resale value where a secondary market exists, the share of sales taken at full price, and what the first page of search results says about you. The full price share is the most honest single number in the category, and the one most often left out of the report. Related. The name carries the brand: see brand naming (https://colleviate.com/blog/brand-naming.html). ## Questions **Can a luxury brand run performance marketing?** Yes, for reach and for efficient retargeting within limits. The damage comes from discount ladders and urgency tactics, which trade long term pricing power for short term volume. **How important is production quality in luxury?** It is the argument, not the wrapper. In a category where the buyer cannot test the craft before purchase, the imagery and the film are the evidence. **Is discounting always wrong in premium?** Not always, but it should be structural and rare: private, time bound, or tied to a genuine reason. Routine discounting converts a luxury brand into an expensive one. --- # Marketing a firm that sells judgement URL: https://colleviate.com/blog/professional-services-marketing.html Topic: Growth audits | Published: 2026-09-19 | Updated: 2026-09-19 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Buyers of professional services are choosing a person more than a firm, so the marketing job is making that expertise visible and checkable. Published thinking, partner visibility and a systematic referral habit outperform advertising, which several professions also restrict. Nobody buys an accountant from an advertisement. They buy one because somebody they trust said a name, and then the website confirmed or ruined it. Professional services sell something the buyer cannot evaluate in advance and will only find out about after committing. In that situation people fall back on proxies: who else uses them, what have they published, who do I know who has worked with them. ## The regulatory frame Several professions in India restrict advertising and solicitation, with the rules for advocates being the most restrictive and chartered accountants also operating under institute guidance. The practical effect is that the permitted channels are informational and reputational rather than promotional, which happens to be what works anyway. ## Published thinking is the product sample It is the only way a prospect can assess judgement before buying it. One substantial piece a quarter, written by the partner whose name is on it, on a question clients actually face, beats a weekly blog of summarised news. It also circulates inside a client organisation when the person who found it needs to convince somebody else, which is the real mechanism: see content that earns links (https://colleviate.com/blog/content-that-earns-links.html). ## Partner visibility is the channel Clients hire people. A firm whose partners are invisible is asking to be selected on price. Speaking, writing under their own names, and being quoted in trade press all do more than any firm-level campaign. ## Referral, as a system Most firms get most of their work from referral and manage none of it. Keep a list of the people who send work, know what they need, stay in contact when you want nothing, and tell them specifically what a good referral looks like. That last part is where most referral flow is lost: people cannot refer what they cannot describe. ## What the website is for Confirming a recommendation. Someone has heard your name and is checking. That means people, credentials, sectors, published work and an obvious way to make contact. Not a carousel, not a mission statement, and not stock photography of a handshake. ## Questions **Can law firms advertise in India?** Advocates operate under strict restrictions on advertising and solicitation, and other professions have their own guidance. Informational and reputational routes are both permitted and more effective. **What is the most effective marketing for a professional firm?** Published thinking under named partners, visible expertise, and a systematically managed referral network. The website exists mainly to confirm a recommendation. --- # How to choose a digital marketing agency in India URL: https://colleviate.com/blog/how-to-choose-a-digital-marketing-agency.html Topic: Method | Published: 2026-09-18 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Choose a digital marketing agency by deciding what you are buying first: a media buying shop, a creative studio, an SEO and content operation, or a full-service agency answerable for the result. Then ask questions that are hard to answer from a deck, read the scope rather than the price, and pay two agencies for a short scoped piece of thinking instead of free pitches. Most agency selections are decided by the pitch, which is the one part of the relationship you will never experience again. Here is what to look at instead. Every agency pitch is a showreel of the work that went well, presented by the people least likely to do your work. That is not dishonesty, it is the format. The buyer’s job is to get underneath it, and the way to do that is to ask questions that are awkward to answer from a deck. ## Decide what you are actually buying “Digital marketing agency” covers at least four different businesses: a media buying shop, a creative studio, an SEO and content operation, and a full-service agency that does all of it and answers for the result. The first mistake is hiring one and expecting another. If your problem is that nobody knows who you are, a performance agency will spend efficiently against a demand that does not exist yet. If your problem is that people know you and still do not buy, a creative relaunch will not fix a broken funnel. Diagnose before you shop. A structured growth audit (https://colleviate.com/blog/growth-audit-what-it-should-contain.html) is the cheapest way to find out which problem you have. ## Seven questions that are hard to fake - Who does the work, and what else are they on? Ask for names and the number of other accounts each one carries. Pitch teams and delivery teams are rarely the same people. - Show me a campaign that failed and what you changed. An agency that cannot describe a failure in specific terms has either not been trusted with anything difficult or is not being straight with you. - What would you refuse to do? Agencies that will do anything usually have no point of view, and a point of view is most of what you are paying for. - How do you decide when to stop a campaign? Listen for a number and a date, not a philosophy. - What does month one look like? Onboarding, access, analytics, a baseline. If month one is “ideation”, budget for a slow start. - Which of your clients can I call, unaccompanied? The word doing the work is “unaccompanied”. - What are you measuring in ninety days, and what is out of your control? An agency that claims everything is in its control is either inexperienced or selling. ## Read the scope, not the pricing Two proposals at the same monthly fee can differ by a factor of three in what they deliver. Compare the number of deliverables, the revision rounds, who owns the assets and the accounts, what happens to your ad account if you leave, and whether media spend is billed through the agency or your own card. The last one matters more than most buyers realise: if the agency holds the account, your history leaves with them. On what the fee itself should look like, see what a marketing agency should cost (https://colleviate.com/blog/what-a-marketing-agency-should-cost.html). ## The retainer trap A retainer is right when the work is genuinely continuous: always-on media, content, SEO, community. It is wrong when you are buying a one-off, like a website or a launch film, because the fee keeps arriving after the work has finished. Mixing the two in a single number is how relationships drift into paying for a monthly meeting. ## Pitch process, without the theatre Do not ask four agencies for free strategy. The work you get is thin by necessity and the agency that invests most in a free pitch is often the one with the least billable work. Instead: pay two agencies a small fee for a scoped piece of thinking, give them the same brief and the same data, and judge the thinking rather than the animation. Our own view is in how to brief a performance marketing agency (https://colleviate.com/blog/how-to-brief-a-performance-marketing-agency.html). ## What a good first ninety days looks like Access and tracking fixed in week one. A baseline nobody argues with by week two. A written diagnosis, including what is not worth doing, by week four. First work live in the first month, not the third. And a monthly review that leads with what did not work, because that is the part with the learning in it. If an agency cannot show you that shape, you are buying activity rather than progress. Related. Most shortlists are formed before the first call: how B2B buyers shortlist agencies before they contact one (https://colleviate.com/blog/how-b2b-buyers-shortlist-agencies.html). ## Questions **How much does a digital marketing agency cost in India?** Retainers commonly run from about Rs 1.5 lakh a month for a narrow scope to Rs 10 lakh and above for full-service work with production. The fee should be judged against the scope and the seniority of the people on it, not compared like for like. **Should I hire one agency or several specialists?** Several specialists give you depth and a coordination problem. One full-service agency gives you coherence and a dependency. Below roughly Rs 5 lakh a month of total marketing spend, coordination usually costs more than the depth is worth. **How long before an agency shows results?** Paid media can move within weeks. SEO, content and brand work take two to three quarters before the trend is readable. Any agency promising fast movement on the slow levers is describing an outcome it cannot control. --- # What to do in the first six hours URL: https://colleviate.com/blog/crisis-communications.html Topic: Culture | Published: 2026-09-18 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** In the first six hours of a crisis, acknowledge quickly even without full facts. The holding statement says three things: we know, we are looking into it, and we will update you by a stated time. Then meet that time. Most reputational damage comes from the response, so prepare the plan and the people on a quiet day. Most reputational damage is done by the response, not by the event. The response is decided by whether anyone prepared for it, which is work nobody wants to do on a good day. Something goes wrong. A product fails publicly, a customer posts a video, an employee says something, a payment breaks, a journalist calls. The event is usually survivable. What is often not survivable is six hours of silence followed by a defensive statement written by a committee. ## Prepare on a quiet Tuesday - A named decision maker, reachable, authorised to approve a statement without a board meeting. - A holding statement template that can be filled in within fifteen minutes. - A list of what could plausibly go wrong, written by the operations team rather than by marketing, with a position drafted for the top five. - Monitoring that would actually tell you, including the platforms where your customers complain rather than the ones you post on. - Agreement on who speaks, and an instruction to everybody else that they do not. This takes a day and it is the entire difference between a controlled response and an improvised one. ## The first six hours Acknowledge quickly, even without full facts. The holding statement says three things: we know, we are looking into it, we will come back by a stated time. Then meet that time, even if the update is that you do not yet know more. Say what you do know and nothing you do not. Do not speculate about cause. Do not blame a vendor, an employee or a customer in public, ever, because it reads as evasion even when it is true. Correct factual errors in coverage, politely and privately first. Fix the thing before talking about having fixed it. A statement that outruns the remedy produces a second cycle worse than the first. ## When silence is correct When the matter is in a legal process, when responding would amplify something almost nobody has seen, or when a single angry individual is seeking engagement rather than resolution. Silence is a decision that should be made deliberately and reviewed every few hours, not a default produced by nobody feeling able to act. ## Tone Plain language, no corporate abstraction. People can tell the difference between an apology and a sentence constructed to avoid one. If you got it wrong, say so in short words. Regret for an outcome without ownership of the cause is the most common tonal failure in Indian corporate statements, and the audience reads it accurately every time. ## Afterwards The incident becomes a search result, and the search result lasts far longer than the news cycle. Publishing your own clear account of what happened and what changed gives both people and answer engines something accurate and first party to find. Then do the slow corroboration work: coverage that is not about the incident, content that is genuinely useful, reviews answered properly. That is reputation repair, and it runs in months rather than weeks: see reputation is a search result (https://colleviate.com/blog/reputation-is-a-search-result.html). ## Questions **How fast should a brand respond to a crisis?** Acknowledge within hours, even without full facts. A holding statement that says what you know and when you will update next buys the time you need and costs nothing. **Should we ever stay silent?** Yes, when the matter is in legal process, when responding would amplify something few people have seen, or when the complainant wants engagement rather than resolution. Make it a reviewed decision, not a default. **How do you repair a reputation after an incident?** Publish your own accurate account, fix the underlying cause, and then do months of corroboration: unrelated credible coverage, genuinely useful content, and reviews answered properly. --- # Marketing for galleries, festivals and cultural institutions URL: https://colleviate.com/blog/arts-and-culture-marketing.html Topic: Culture | Published: 2026-09-18 | Updated: 2026-09-18 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Cultural marketing has two jobs that pull against each other: selling this programme and building an audience for the next decade. Budget them separately, because the short-term ticket push will otherwise eat the long-term work every single time. The work is extraordinary and the marketing is a poster two weeks out. This is almost universal and almost always a resourcing problem rather than a taste one. Cultural organisations are selling an experience nobody can preview, often by an artist the audience has not heard of, against the easiest alternative in the world: staying home. ## Two budgets, not one Selling tickets to the current show is demand capture. Building an audience that comes to whatever you programme is demand creation. They use different channels, different timelines and different measures, and merged into one budget the urgent one wins permanently. Protect the audience work in writing, because the fortnight before opening will always feel more important than the decade. ## The programme is the marketing What you choose to show, who you invite, how you contextualise it. A programme that gives somebody a reason to tell a friend is doing more than any campaign, and in this category word of mouth is not a nice-to-have, it is the primary channel. ## Access is a marketing decision Price, timing, language, physical access, and how intimidating the building feels from the pavement. Many institutions in India have an audience problem that is really a threshold problem: people who would love the work believe it is not for them. Free hours, student pricing and visible welcome do more for attendance than a wider media buy. ## Document everything Photograph and film the work, the installation, the audience, the talks. It is the only inventory you will have for the next programme, for funders and for press, and institutions that skip it spend every cycle starting from nothing. ## Funders read differently Attendance numbers matter, but reach, demographic spread, education outcomes and press coverage often matter more in a grant report. Collect those deliberately during the programme rather than reconstructing them afterwards, and keep the earned coverage filed as it happens: see PR worth paying for (https://colleviate.com/blog/pr-worth-paying-for.html). ## Questions **How should cultural organisations split marketing budget?** Separate ticket sales for the current programme from long-term audience development, and protect the second in writing. Otherwise the urgent one consumes it. **What is the biggest barrier to attendance?** Usually the threshold rather than the price or the programme: people who would enjoy the work believe it is not meant for them. --- # What a marketing agency should cost, and what you get for it URL: https://colleviate.com/blog/what-a-marketing-agency-should-cost.html Topic: Method | Published: 2026-09-17 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Agency pricing follows senior hours, the real unit being sold. Retainers buy a share of those hours for continuous work; projects buy a defined outcome. To compare proposals, convert both to hours by seniority, count deliverables rather than adjectives, find the measurement line, and confirm in writing that ad accounts, analytics, domains and files are yours. Agency pricing looks arbitrary from the outside because the unit being sold is rarely stated. It is almost always senior hours, and everything else follows from that. Strip away the packaging and an agency sells attention from people who have done this before. A retainer is a claim on a certain number of those hours each month. Once you see the fee that way, most pricing questions answer themselves. ## The three models Retainer. A fixed monthly fee for a defined scope. Right for continuous work: media management, content, SEO, social, always-on creative. Predictable for both sides, and the easiest to quietly stop earning. Project. A fee for a defined outcome: a brand refresh, a website, a launch campaign, a film. Right when the work has an end. Watch for scope creep, and agree the number of revision rounds in writing. Performance. A fee tied to a result, usually a percentage of media spend or a cost per acquisition. Superficially attractive, structurally awkward: a percentage of spend rewards spending, and a cost-per-lead deal pushes the agency towards the cheapest leads rather than the best ones. Use it as a component, rarely as the whole deal. ## What sits inside a monthly fee For a working Indian mid-market retainer, the split usually looks something like this: strategy and account leadership, creative and production, channel execution, and analytics. If a proposal has no analytics line, you are paying for activity with no way to judge it. If production is a large share, check whether you need that volume or whether it is there to justify the number. Broadly, what we see in the market: narrow single-channel scopes from about Rs 1.5–3 lakh a month; a credible multi-channel scope with real creative from Rs 4–8 lakh; full-service with production and senior time, Rs 10 lakh and up. Film and photography projects price separately, and production costs (https://colleviate.com/blog/video-production-cost-india.html) vary far more than retainers do. ## Why the cheapest quote is usually the most expensive Agencies do not have secret cost advantages. If one quote is half another for the same scope, something has been removed: seniority, time, or the analytics that would have shown it. The usual result is work that is technically delivered and commercially inert, plus a year lost before you find out. The honest version of a low fee is a smaller scope, and a good agency will offer you that instead. ## How to compare two proposals - Convert both to hours. Ask each agency how many people, at what seniority, for how many hours a month. The answer is revealing even when it is vague. - Count deliverables, not adjectives. Twelve assets a month is a number. “Content support” is not. - Find the analytics line. Someone has to own measurement, and it is a real cost. See the measurement gap (https://colleviate.com/blog/marketing-measurement-gap.html) for why it gets cut first and hurts most. - Check ownership. Ad accounts, analytics properties, domains, source files, raw footage. All yours, in writing. - Read the exit clause. Notice period, handover, what you receive on the last day. ## The ratio that matters Agency fee against media spend is the number most finance teams ask about. Below roughly Rs 3 lakh a month of media, a full-service retainer is hard to justify on efficiency grounds alone; you are buying capability you do not yet have, which can still be the right call. Above Rs 15–20 lakh a month, the management fee should fall as a percentage or you should be getting materially more than management. On sizing the whole budget rather than the fee, see how much to spend on marketing (https://colleviate.com/blog/marketing-budget-how-much-to-spend.html). ## What we would push back on Paying for a monthly report that nobody reads. Paying a percentage of spend with no floor on quality. Paying for “strategy” as a separate line every quarter when it should be continuous. And paying for volume of content when the evidence says the problem is distinctiveness, not quantity. Related. Before you compare quotes: how to choose a digital marketing agency (https://colleviate.com/blog/how-to-choose-a-digital-marketing-agency.html). ## Questions **Is a retainer better than a project fee?** Only when the work is genuinely continuous. For anything with an end date, a project fee is cleaner and usually cheaper. Many relationships work best as a small retainer plus projects. **What is a reasonable agency fee as a share of media spend?** Commonly 10 to 20 percent for management-only engagements, falling as spend grows. Where the agency also makes the creative, comparing fee to spend alone understates what is being bought. **Should the agency bill our media spend?** Prefer paying the platforms directly with your own card where you can. You keep the account history, the data and the leverage. --- # Real estate marketing that produces site visits URL: https://colleviate.com/blog/real-estate-marketing-india.html Topic: Growth audits | Published: 2026-09-17 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Real estate marketing should be optimised for site visits, not leads. Price the funnel by cost per lead, per contactable lead, per site visit and per booking; most campaigns look efficient on the first and poor on the third. Qualify in the ad with price, location and configuration, respond fast, and keep creative within RERA rules. Property marketing is measured in leads and decided by site visits. Those two numbers move independently, and most budgets are optimised against the wrong one. Every developer has a lead problem that is actually a qualification problem. The forms fill, the calls go unanswered, and the agency is asked for more leads, which produces more of the same. The constraint is rarely at the top of the funnel. ## Lead quality beats lead volume, measurably Price the funnel properly: cost per lead, cost per contactable lead, cost per site visit, cost per booking. Most campaigns look efficient on the first number and terrible on the third. Once the team sees cost per site visit by source, the budget reallocates itself without an argument. The levers on quality are unglamorous. Qualify in the ad, not after it: show the price band, the location and the configuration. A click that leaves after seeing the price saved you a phone call. ## Speed decides the outcome In this category, response time is the single highest leverage variable available. A lead called back within minutes converts at a multiple of one called the next morning, and property buyers are enquiring on several projects in the same session. If the sales team cannot respond immediately, the spend is partly funding your competitors' conversations. This is an operations fix, not a marketing one, and it is routinely cheaper than the media budget being argued over. ## Portal dependence The listing portals bring volume, charge for it, and sell the same buyer to everyone on the page. They are useful and they should not be the whole plan. The parallel asset is your own project site, with the site plan, the real photographs, the construction status and a price indication, ranking for the searches that name the location and the configuration. Those searches are high intent, winnable, and permanently cheaper than portal leads. The dependence logic is the same as in marketplace or your own store (https://colleviate.com/blog/marketplace-or-own-store.html). ## What moves site visits - Real footage of the actual site, including construction progress. Renders alone read as evasive to anyone who has bought before. - Location content, which is what people genuinely research: schools, commute, infrastructure, what the area will look like in five years. - Price transparency, or at least a band. Hiding it increases leads and reduces visits. - Proof of delivery. Completed projects, handover dates met, and existing residents saying so. ## RERA and the creative Advertising for a registered project carries real obligations: the registration number, accurate carpet area, no promises the approvals do not support. Treat this as a creative constraint rather than a legal afterthought. The brands that handle it well make transparency part of the pitch, which in a category with a trust deficit is a genuine differentiator rather than a compliance cost. ## The measurement that matters Report cost per site visit and cost per booking by source, with the lag acknowledged, since the booking may be ninety days after the click. Track the drop between enquiry and contact, which is where most of the loss sits. And look at how many visits each booking takes, because that number tells you whether the problem is the marketing or the project. ## Questions **How quickly should real estate leads be called?** Within minutes. Buyers enquire across several projects in one session, and response time is usually a bigger lever than media spend. **Should developers rely on listing portals?** They are useful for volume and should not be the whole plan. Your own project pages ranking for location and configuration searches produce cheaper, better qualified enquiries over time. **Does showing the price reduce leads?** Yes, and it raises site visits, which is the number that matters. Hiding price inflates the top of the funnel with people who will never transact. --- # Travel marketing between inspiration and booking URL: https://colleviate.com/blog/travel-and-tourism-marketing.html Topic: Growth audits | Published: 2026-09-17 | Updated: 2026-09-17 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Travel decisions have a long inspiration phase and a short booking window. Win the first with content and social proof, win the second with availability, price clarity and a booking path with no friction, and accept that attribution across a four-month gap will always be partial. Somebody decides to go somewhere months before they book it, and the brand that shaped the decision is rarely the one that gets the credit. The category splits cleanly into two problems that need different work: making somebody want to go, and making it easy once they have decided. ## The dreaming phase Long, unstructured, and dominated by other people's photographs. This is where social, creators and genuinely useful content earn their place, and where direct response measurement will tell you almost nothing. The honest measure here is branded search and direct traffic over months, not attributed bookings this week. ## The booking window Short, decisive and ruthless about friction. Price visible, availability accurate, cancellation terms plain, payment options that work locally, and a path that survives a phone on patchy data. Most direct booking losses happen here rather than in the choice of property. ## Platform dependence Booking platforms bring volume at a commission that hurts and own the customer afterwards. The workable position is to use them for discovery while making the direct route genuinely better: a rate advantage where parity rules allow, a guaranteed upgrade, flexibility, something concrete. Parity with nothing extra does not move anybody, which is the same argument as in hospitality and restaurant marketing (https://colleviate.com/blog/hospitality-and-restaurant-marketing.html). ## Seasonality is the planning problem Demand arrives in waves set by holidays, weather and school calendars, and the marketing has to run months ahead of each one. Off-peak needs a different proposition rather than a discount: a reason the shoulder season is actually better, which is frequently true and rarely argued. ## Guests make better content than you do Ask, make it easy, get permission in writing, and use it. Guest photography outperforms produced imagery on trust and costs a fraction. Keep produced work for the things guests cannot shoot: the room at the right hour, the food properly lit, the drone shot nobody is allowed to fly. ## Questions **How do hotels reduce dependence on booking platforms?** Give the direct channel a concrete advantage and remove friction from the booking path. Use platforms for discovery and off-peak fill rather than for repeat guests. **How do you attribute a travel booking made months after the inspiration?** Imperfectly. Use branded search and direct traffic trends for the inspiration phase and last-click only for the booking window, and do not let the attributed number set the whole budget. --- # How much should you spend on marketing? URL: https://colleviate.com/blog/marketing-budget-how-much-to-spend.html Topic: Measurement | Published: 2026-09-16 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Across most industries marketing budgets sit between 5 and 12 percent of revenue, lower in B2B and higher in consumer; early D2C brands often spend far more. Use the benchmark as a sanity check, not a target. The split between demand creation and capture, and whether anyone can tell what the spend did, matter more than the percentage. The honest answer is that the percentage matters less than the split, and the split matters less than whether anyone can tell what the spend did. Every budget conversation starts with a benchmark and ends with a negotiation. The benchmarks are worth knowing, mostly so you can stop arguing about them and get to the questions that actually decide the outcome. ## The benchmarks, briefly Across most industries, marketing budgets sit somewhere between 5 and 12 percent of revenue, with B2B at the lower end and consumer at the higher. Early D2C brands routinely run far above that, sometimes 25 percent or more, because they are buying a customer base rather than defending one. Mature B2B companies with strong referral flow can hold below 5 percent for years and be perfectly healthy. Use the benchmark as a sanity check, not a target. A number that is defensible against your own pipeline maths will survive scrutiny better than one borrowed from a survey of companies you do not resemble. ## The split that decides the outcome The more consequential decision is how much goes to demand capture and how much to demand creation. Capture is search, retargeting, marketplaces: harvesting intent that already exists. Creation is brand, content, PR, social: making intent exist in the first place. Capture-only budgets look excellent for about eighteen months and then plateau, because you cannot harvest more demand than the market generates. Creation-only budgets look wasteful for about the same period and then start showing up in branded search and in the cost of everything else. Our longer argument is in brand versus performance (https://colleviate.com/blog/brand-versus-performance-split.html). ## What to fund first, by stage - Pre-product-market-fit. Fund learning: a small search budget, one credible content surface, and analytics you trust. Do not fund awareness for a proposition still being rewritten. - Early traction. Fund capture hard, because intent already exists and is cheap to serve. Keep a fifth aside for creation so the plateau arrives later. - Scaling. Shift towards creation deliberately. The signal to watch is capture costs rising while volume stalls. - Mature. Defend distinctiveness, fund retention and category coverage. This is where budget quietly drifts into maintaining reporting rather than buying anything. ## The part that gets underfunded every time Measurement. It is not glamorous, it produces no assets, and it is the first line cut when the number needs to come down. It is also the only reason anyone can tell whether the other 95 percent worked. Budget for it explicitly, including the analytics implementation and someone’s time to interpret it. See the measurement gap (https://colleviate.com/blog/marketing-measurement-gap.html). ## How to defend the number Boards rarely reject a marketing budget because it is large. They reject it because it is unaccountable. Three things make it defensible: - Name what each tranche buys. This much buys pipeline this quarter, this much buys demand next year, this much buys the ability to tell the difference. - State what you will stop. A budget with no kill criteria reads as a wish. - Show a leading indicator. Pipeline lags. Share of search (https://colleviate.com/blog/share-of-search-leading-indicator.html) moves earlier and is cheap to track, which makes it useful evidence between board meetings. ## The cut that does the least damage When the budget must fall, cut breadth before depth. Fewer channels done properly beats every channel done thinly, because thin presence in six places produces no memory anywhere. Protect the work that compounds — distinctive assets, owned content, search position — and cut the work that stops the day you stop paying for it. That order is unpopular, because the compounding work is the least visible this quarter, and it is still the right order. ## What do companies actually spend? The most quoted benchmark is Gartner's annual CMO Spend Survey. Its 2025 edition, a survey of 402 marketing leaders in North America, the UK and Europe at mostly large companies, found marketing budgets flat at 7.7 percent of overall company revenue, the same as the year before. The average hides a wide spread: half the respondents reported budgets of 6 percent or less. Paid media took the largest share, about 30.6 percent of the total marketing budget, and 59 percent of CMOs said they did not have enough budget to execute their strategy. Two cautions before using that number. It describes large companies in mature markets, so a young Indian D2C brand buying its first customers will rightly spend a far higher share, and a long-established B2B firm with strong referrals can spend less. And a percentage of revenue says nothing about the split, which is where most of the outcome is decided. ## How to build the number from the bottom up A percentage of revenue is a check, not a method. The defensible budget is built from the task. Start from the target, say a number of new customers, and work backwards through the conversion rates you can evidence to the reach and demand you need. Price the demand you can capture from search and existing channels first, because it is the cheapest. Price the demand you will have to create next, because it is the slowest. Then add the fixed costs people forget: creative, production, tools and measurement. Compare the total with the benchmark for your stage. If yours is far above it, the plan is probably asking marketing to fix a product or pricing problem. If it is far below, the target is probably a wish. Either way the conversation with the finance director is about the assumptions, which is the conversation you want to be having. ## Sources - Gartner 2025 CMO Spend Survey reveals marketing budgets have flatlined at 7.7% of overall company revenue, Gartner, 12 May 2025: https://www.gartner.com/en/newsroom/press-releases/2025-05-12-gartner-2025-cmo-spend-survey-reveals-marketing-budgets-have-flatlined-at-seven-percent-of-overall-company-revenue ## Questions **What percentage of revenue should go to marketing?** Commonly 5 to 12 percent, lower for B2B and higher for consumer brands. Early-stage D2C often runs much higher because it is buying a customer base rather than defending one. **How should we split brand and performance spend?** A long-run split near 60 brand to 40 performance is the widely cited benchmark for established consumer brands. Younger companies reasonably invert it, then shift as capture costs start rising faster than volume. **What should we cut first when budgets tighten?** Breadth, not depth. Drop the channels you are present in thinly and protect the work that compounds: search position, distinctive assets, owned content. --- # Industrial marketing for companies that sell to engineers URL: https://colleviate.com/blog/manufacturing-and-industrial-marketing.html Topic: Growth audits | Published: 2026-09-16 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Technical buyers shortlist from specification documents, not from brand messaging. Publish complete, downloadable technical data, make the catalogue searchable, and treat the datasheet as the most important marketing asset the company owns. An engineer evaluating a component wants a datasheet. Most manufacturer websites offer a brochure and a contact form. Industrial buying is specification-led. Somebody has a requirement, searches for a part or a capability that meets it, and builds a shortlist from whoever published enough detail to be evaluated. Brand sentiment barely enters the first round. ## Publish the specifications Complete technical data, dimensions, tolerances, materials, certifications, compatibility, downloadable without a form. Gating a datasheet behind a lead capture form removes you from the shortlist, because the engineer simply uses a competitor who did not. The lead you collected is worth less than the specification round you lost. ## Make the catalogue findable Every product needs its own indexable page with the part number, the specifications and the applications. Catalogues locked inside a PDF are invisible to search, which means the long tail of exact-part queries, the highest-intent traffic in the category, goes to distributors instead of to you. ## Distributor conflict, handled Generating demand that your distributors fulfil is a feature, not a conflict, provided the handover is explicit. Decide what you own, state where to buy, and give distributors the material to close with. Competing with your own channel quietly is worse than either selling direct or not. ## Trade shows still work here In few categories does this remain true, and industrial is one of them, because the buyer wants to handle the thing and meet the engineer. The return comes from pre-show outreach and post-show follow-up rather than from the stand, and most exhibitors budget for the stand and not for either. ## Content that gets you shortlisted Application notes, selection guides, failure analysis, comparison tables against alternative approaches, and installation documentation. It reads as unglamorous and it is exactly what the buyer searches at the moment they are deciding, which is the argument in content that earns links (https://colleviate.com/blog/content-that-earns-links.html). Related. Five measures for the channel your dashboard cannot see: measuring a business that sells through others (https://colleviate.com/blog/measuring-a-business-that-sells-through-others.html). ## Questions **Should technical datasheets be gated behind a form?** No. Engineers evaluating options will use a supplier who published openly, so gating removes you from the shortlist in exchange for a low-value lead. **Do trade shows still deliver for manufacturers?** More than in most categories, because buyers want to handle the product. The return comes from pre-show outreach and follow-up, not from the stand itself. --- # The brief is not the idea URL: https://colleviate.com/blog/the-brief-is-not-the-idea.html Topic: Creative | Published: 2026-09-15 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A working creative brief is a one-page constraint that provokes a thought, not a document written to survive review. It is killed by vague adjectives, generic tone-of-voice paragraphs, mandatories longer than the problem, and objectives that are really channel plans. Evidence belongs in the brief; it sharpens the idea rather than shrinking it. Every agency has a brief template. Almost none of them have a rule for when the brief is finished. So briefs get longer, tighter, more thoroughly approved, and the idea gets smaller. A brief is a constraint, and constraints make ideas. That is the theory. In practice most briefs are written to survive a review rather than to provoke a thought, and there is a difference you can feel in the room. ## What a working brief actually contains Four things, and they fit on a page. - The tension. Not the audience profile. The specific thing a real person believes that stands between them and the brand. - What we know. The two or three pieces of evidence that make the tension credible rather than asserted. Where the demand actually sits, what the category already says, what people write when nobody is asking. - The one job. What this piece of work has to do that nothing else is doing. - What would make it wrong. The result that tells you the idea missed, agreed before anyone falls in love with it. ## What kills a brief Three adjectives where one noun would do. A tone-of-voice paragraph that could belong to any brand in the category. A list of mandatories longer than the problem statement. And the most common one: an objective that is really a channel plan wearing a strategy costume. ## Evidence is not the enemy of the idea There is a persistent belief that data and creative pull in opposite directions, that the more you know the less you dare. We have not found that to be true. What kills daring is ambiguity, not evidence. When you genuinely know where demand is coming from, what the category sounds like, and which belief is doing the blocking, the creative team can stop hedging and commit. The useful test is whether the evidence in the brief narrows the field or just decorates it. A chart nobody argues with is decoration. A finding that rules out three obvious routes is a brief. ## How we write them Short. One page, four headings, a named tension, and a line that says what would prove us wrong. Then the creative team is free to ignore the format entirely, which is the point of having one. Related. The practical side is how to write a brief (https://colleviate.com/blog/writing-a-brief.html). ## Questions **How long should a creative brief be?** One page. If the tension cannot be stated in a sentence, the thinking is not finished, and adding pages will not finish it. **Should the brief include the data?** Only the two or three findings that change what the work should be. Evidence that does not rule anything out is decoration. --- # What a campaign owes the culture it borrows from URL: https://colleviate.com/blog/what-a-campaign-owes-the-culture.html Topic: Culture | Published: 2026-09-15 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A campaign that borrows from a culture owes it accountability: understanding, fair payment and credit to the people inside it. Audiences can tell within seconds whether a brand understands a subculture or is borrowing a mood board. Real participation is slower and costs more, and it earns trust that borrowed references spend. Every brand wants to be in the conversation. Very few want to be accountable to it. The gap between those two is where most cultural marketing quietly fails. Culture is the cheapest raw material in marketing. A sound, a phrase, a visual language, a movement, all of it sitting there apparently free to use. The bill arrives later, and it is usually paid in trust. ## Three questions before you borrow - Did anyone from inside make this with you? Not consulted at the end. Involved when it could still change. - Would this still make sense if the brand were removed? If the work only holds together as an ad, the culture is a costume. - What happens the day after? A campaign that arrives, extracts attention, and leaves has told the community exactly what it was worth. ## Audiences read the difference instantly People who live inside a subculture can tell whether a brand understands it within seconds, usually from small details: the wrong slang register, a reference two years stale, casting that looks like a mood board rather than a room. They are not being harsh. They have simply seen it done badly many more times than you have done it at all. ## Participation costs more and is worth more Real participation is slower. It means paying people properly, crediting them publicly, giving up some control of the output, and sometimes being told your idea is wrong by somebody who knows better. It also produces work that survives contact with the audience, which the extractive version never does. ## The commercial case, since you will be asked for one Cultural credibility compounds and cultural embarrassment is permanent. A campaign that lands inside a community earns a second hearing next season. One that gets it wrong makes every subsequent attempt harder, and the internet does not forget the screenshot. Related. A cultural history of advertising that did this well and badly: iconic Indian ads for the ages (https://colleviate.com/blog/iconic-indian-ads-for-the-ages.html). ## Questions **How do you know if a cultural reference has gone stale?** Ask the people who use it, not a trends deck. By the time a reference reaches a trends deck it has usually been over for months. **Does this apply to local Indian culture too?** Especially. Regional language, festival and city-specific culture gets flattened into a generic national idea more often than almost anything else. --- # Casting for evidence, not for follower count URL: https://colleviate.com/blog/casting-for-evidence.html Topic: Creative | Published: 2026-09-15 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Cast creators on evidence about who actually watches them, not on follower count. The largest creator in a category is usually the most expensive and most crowded, with an audience that has seen every competitor. Good casting is usually several mid-sized creators with genuine authority plus one wildcard, briefed, credited and paid properly, then measured honestly. Follower count is the easiest number in the brief to defend and the least connected to whether anything sells. It survives because it is simple, not because it works. The creator economy has produced a strange procurement habit: brands buy audience size the way you would buy media impressions, then act surprised when the response looks nothing like the reach. ## What we look at instead - Comment substance. Not volume. Whether people are discussing the thing or performing engagement. Thirty real questions beat three thousand fire emojis. - Repeat audience. Whether the same people keep coming back, or each post finds a new crowd that never returns. - Category adjacency. Whether they already talk about the problem your product solves, before anyone paid them to. - Search pull. Whether their name generates branded search, which is the cleanest signal that attention converted into intent. - Previous brand work. How many partnerships in the last six months, and whether their audience visibly tired of them. ## The big account problem The largest creator in a category is usually the most expensive and the most crowded. Their audience has seen every competitor pass through. What you buy is reach into a room where your category has already been discounted. ## What good casting looks like Usually several mid-sized creators with genuine authority and one wildcard from outside the category who will make the work interesting. Briefed properly, credited properly, paid properly, and given enough latitude that the post sounds like them rather than like your brand guidelines read aloud. ## Then measure it honestly Branded search lift, direct traffic, code redemption, and the shape of the comment section. Not reach. Reach is what you bought, not what you got. Related. The wider programme: influencer marketing that works (https://colleviate.com/blog/influencer-marketing-that-works.html). ## Questions **How many creators should a campaign use?** Enough to see a pattern and few enough to brief properly. For most brands that is five to twelve, not fifty. **What is the single best signal a creator will work?** Whether their audience already discusses your category unprompted. Everything else is easier to fake. --- # Reputation is a search result URL: https://colleviate.com/blog/reputation-is-a-search-result.html Topic: Culture | Published: 2026-09-15 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** For most brands, reputation is now what comes back when someone searches the name: a results page and increasingly an AI-generated summary. Assistants weight consistency and corroboration, so the fix for an unflattering answer is to make the accurate fact appear in several sources the model already trusts, and to resolve the inconsistency causing it. For most brands, reputation is no longer a conversation. It is a page of results and a paragraph generated by an assistant, and both are assembled from sources you may never have looked at. Somebody considering you will type your name. What comes back is your reputation, whatever your brand tracker says. And increasingly what comes back is not a list at all, but a synthesised paragraph that has already decided which sources were worth believing. ## The three surfaces - The results page. Ten links, a knowledge panel, reviews, and whatever the category thinks of you. - The answer. An assistant summarising you in a paragraph, drawing from sources it considers consistent and corroborated. - The rooms you cannot see. Group chats, forums, comment sections. Unindexed, uninfluenceable, and usually the origin of everything that surfaces later. ## Why the second one changes the work A results page can be improved by adding good pages. An answer cannot, not directly. Assistants weight consistency and corroboration, so the fix for an unflattering summary is rarely to publish another page. It is to make the accurate version of the fact appear in several places the model already trusts, and to resolve whatever is causing the inconsistency. ## What actually helps - Fix the factual record first: your own pages, your listings, the directories and reference sources assistants lean on. - Earn coverage in places with editorial weight, not just volume. One credible outlet outranks twenty syndicated reprints. - Answer reviews like a person, in public, including the fair criticism. It reads better to humans and it adds corroborated context for machines. - Track the answer, not only the ranking. Ask the assistants the questions your buyers ask, monthly, and record what comes back. ## What does not help Burying a bad result under thin content. Review-gating. Threatening a reviewer. All three are visible, all three eventually become the story, and the last one has a habit of arriving with a screenshot. Related. The profiles that describe you should agree with your site; Organization schema and sameAs (https://colleviate.com/blog/organization-schema-and-sameas-for-agencies.html) shows how to link them. ## Questions **How long does reputation work take?** Months, not weeks. The factual record can be corrected quickly; the corroboration that changes an AI answer accumulates slowly. **Can a negative result be removed?** Rarely, and usually only where it breaches the platform policy or the law. The practical work is making the accurate version more findable and better corroborated. --- # Audacity gets you noticed. Tenacity gets you chosen. URL: https://colleviate.com/blog/audacity-and-tenacity.html Topic: Creative | Published: 2026-09-15 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Audacity and tenacity are two different currencies. Audacity, the brave campaign, buys entry: it is the only reliable way to get noticed, but it is expensive and short-lived. Tenacity, the same idea held for years, buys memory, because distinctiveness comes from repetition far more than novelty. Brands that compound spend both deliberately, in the right order. Nearly every brand has a bias. Some are brave once a year and forgettable in between. Others are relentlessly consistent about something nobody noticed in the first place. Both feel like strategy from the inside. Audacity is the swing: the campaign that risks being disliked, the idea nobody has permission to approve, the line that makes the legal team ask for a call. Tenacity is the unglamorous part: the same idea, held for years, through three marketing directors and a budget cut. The brands that compound are not the brave ones or the consistent ones. They are the ones that understand these are two different currencies, and that spending one does not earn you the other. ## What each one actually buys Audacity buys entry. It is the only reliable way to get noticed in a category where everyone has the same media plan and roughly the same product. It is expensive, it is risky, and it has a short half-life. A brave campaign that is not followed by anything is a firework: genuinely impressive, entirely gone by Tuesday. Tenacity buys memory. Distinctiveness comes from repetition far more than from novelty, which is inconvenient for anyone whose job is to be interesting. The asset only becomes an asset on the hundredth exposure. Most brands abandon theirs at the twentieth, usually because the people who made it are bored of it, which is the single worst reason to change anything. ## The two failure modes All audacity. A reel of brilliant one-offs with nothing connecting them. Great for awards, strange for the buyer, who has now met four different versions of you. You can spot this brand by its internal language: every year is a relaunch, every brief starts with "we need to do something different." All tenacity. Immaculate consistency in service of a position nobody has ever repeated back to you. Fifteen years of the same typeface and the same promise, correctly applied, entirely unremarkable. This brand talks about brand guidelines more than it talks about the market. ## How to tell which one you are short of This is an evidence question, not a taste question, and it is answerable. - Ask people to describe you without prompting. If the descriptions are vivid but contradictory, you have audacity and no tenacity. If they are consistent but generic, the reverse. - Look at branded search over three years. Spikes with no floor lift mean you keep buying attention and giving it back. A rising floor means something is sticking. - Count how many campaign ideas you have run in five years. More than five is usually churn, not courage. - Check whether your distinctive assets survive a squint test. If the logo has to be visible for people to know it is you, you have no assets, only a logo. ## Spending them in the right order The sequence matters. Audacity without a plan to sustain it wastes the attention it wins, so the brave move should be the thing you intend to keep doing, not a detour from it. The useful discipline is to decide, before the work is made, which parts are permanent and which are this year’s expression of them. In practice that means a small number of things you will not change — the position, one or two assets, the tone — and generous freedom everywhere else. Tenacity at the level of meaning, audacity at the level of execution. Brands that invert this produce the worst of both: a wildly different promise every year, delivered in exactly the same layout. ## The uncomfortable part Tenacity is mostly a governance problem, not a creative one. The idea rarely dies because it stopped working. It dies because a new stakeholder needs a mark on the wall, or because an agency needs something fresh to present, or because the quarterly review rewards novelty over compounding. Which is why we argue for deciding it in writing, at the start: what is permanent, what is seasonal, what would have to be true for us to change the permanent things. Then when somebody wants to relaunch in month fourteen, the conversation is about evidence rather than appetite. ## Both, deliberately Be audacious about the idea. Be tenacious about holding it. And be honest about which one your brand has actually been spending, because almost nobody is short of opinions on this and almost everybody is wrong about themselves. ## Questions **How long should a brand hold a campaign idea?** Longer than it feels comfortable. Most ideas are retired around the point the market is beginning to recognise them, which is the worst possible moment. **Is audacity riskier than consistency?** In the short term, yes. Over five years the bigger risk is spending consistently on something nobody noticed, because that failure is invisible until it is expensive. **How do you get a board to back a bold idea?** Show what the evidence rules out. A brave route is much easier to approve when the safe routes have been demonstrated not to work. --- # Hospitality and restaurant marketing URL: https://colleviate.com/blog/hospitality-and-restaurant-marketing.html Topic: Growth audits | Published: 2026-09-15 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Hospitality marketing is the business of producing credible evidence for an experience nobody can sample before paying. Use aggregators and travel sites for discovery while growing direct business, treat reviews as the distribution channel, make photography do the work of the menu, and keep local search listings complete and accurate. In hospitality the product is an experience that nobody can sample before paying, so the marketing is mostly the business of producing credible evidence. Hotels and restaurants share a structural problem. The decision is made on a phone, in a few minutes, against several alternatives, by somebody who cannot taste the food or see the room. Everything that follows is about reducing that uncertainty faster than the competition. ## The aggregator question Delivery platforms and travel sites bring volume at a commission that would be unacceptable in most other industries. They are also where the demand is, which makes righteous independence an expensive position. The workable stance is to use them for discovery and to work continuously on the share of business that arrives directly. For restaurants, that means your own ordering route for repeat customers, who are the profitable ones. For hotels, it means a direct booking that is genuinely better: a rate parity exception where allowed, a room guarantee, a late checkout, something concrete. A direct channel that offers nothing extra does not win, it just exists. ## Reviews are the distribution channel Ratings decide placement on aggregators and in map results, and placement decides volume. That makes review management an operational function rather than a marketing one. Ask every satisfied guest, reply to everything including the harsh ones, and treat a pattern in complaints as product feedback rather than as a reputational problem to be managed. The reply is read by future customers more than by the reviewer. A defensive reply costs more bookings than the review did. See reputation is a search result (https://colleviate.com/blog/reputation-is-a-search-result.html). ## The photography is the menu In this category the imagery is not decoration, it is the product description. Dishes shot properly, rooms shot at the right time of day, the room as it actually is rather than a wide angle fiction that generates a disappointed review on arrival. Phone photography by the floor manager is visible, and it is visible next to a competitor who paid for a shoot. Short video outperforms stills for both categories, because motion conveys atmosphere, which is what is actually being sold. ## Local search does the quiet work Most restaurant discovery and a large share of hotel discovery begins in maps and local search. A complete profile, accurate hours, current photographs, menu or rates, and a steady flow of recent reviews will out-perform most social budgets in the same period. The details are in local SEO for service businesses (https://colleviate.com/blog/local-seo-for-service-businesses.html). ## What to measure For restaurants: covers, repeat rate, direct versus aggregator share, average order value, and the review score trend by location. For hotels: direct booking share, revenue per available room, review score, and the cost of acquiring a direct booking against the commission avoided. Not followers. A restaurant with forty thousand followers and empty tables on a Wednesday has an audience problem that is not a marketing problem, and the follower number will keep rising while the business declines. ## Questions **How do we increase direct bookings?** Give the direct channel a concrete advantage, such as a guarantee, an upgrade or a late checkout, and make the booking faster than the aggregator. Parity with nothing extra does not shift behaviour. **Are aggregators worth the commission?** For discovery and for filling off-peak capacity, usually yes. The error is letting them own the repeat customer, who is where the profit is. **How important are reviews for restaurants?** They are the distribution channel. Rating drives placement on aggregators and in map results, and placement drives volume more reliably than any campaign. --- # Fundraising communication that respects everyone involved URL: https://colleviate.com/blog/nonprofit-marketing-india.html Topic: Culture | Published: 2026-09-15 | Updated: 2026-09-15 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Retention beats acquisition in fundraising as decisively as it does in subscription business: a second gift costs a fraction of a first. Report specifically on what the money did, and represent the people you serve as participants rather than as evidence of need. The sector has a technique that works and should mostly not be used, which makes this the one category where the ethics and the strategy genuinely conflict. Distressing imagery raises money. It also strips dignity from the people it depicts, trains donors to give from discomfort rather than conviction, and produces exactly the donor who does not give again. Both halves of that are true, which is why the argument never resolves. ## Retention is the whole economics Acquiring a donor costs many times more than keeping one, and most Indian nonprofits put almost all their communication effort into acquisition. The second gift is the one that makes the first profitable. That means thanking properly and quickly, reporting specifically, and contacting people when you do not want money. The mechanics are the same as any lifecycle programme, as in email and lifecycle marketing (https://colleviate.com/blog/email-and-lifecycle-marketing.html), with higher stakes on tone. ## Report what the money did Not what the organisation did in general. What this donor's contribution, or contributions of that size, changed, with numbers and a timeframe. Vague impact reporting is the most common reason regular donors lapse, and it is entirely self-inflicted. ## Representation Ask whether the person in the photograph would choose this image of themselves. Use names when consent is genuine and informed. Show agency rather than helplessness. The sector has moved on this and Indian fundraising communication still lags, particularly in imagery used for overseas audiences. ## CSR money is a different sale Corporate funding under the CSR regime is a procurement process with compliance requirements, reporting obligations and a corporate counterpart who needs to justify the choice internally. It needs documentation, governance evidence and measurable outcomes, not an emotional appeal. Many organisations use the same materials for both and convert neither well. ## The test Would you be comfortable if the people in your campaign saw every piece of it, including the fundraising email subject lines? That question resolves most of the difficult calls faster than any policy document. ## Questions **Does emotional imagery work in fundraising?** It raises money in the short term and attracts donors who do not give again, while costing the dignity of the people depicted. Agency-led representation retains better. **How is CSR fundraising different from individual giving?** It is a procurement process with compliance and reporting requirements. It needs documentation and measurable outcomes rather than an emotional appeal. --- # Brand versus performance is a false choice, and the split still matters URL: https://colleviate.com/blog/brand-versus-performance-split.html Topic: Measurement | Published: 2026-09-14 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Brand marketing creates demand; performance marketing harvests it. Run either alone and the other's absence shows up as rising costs: a known brand intercepts demand more cheaply and converts better, and brand spend without a capture layer hands demand to competitors. The split matters less as a fixed ratio than as sequencing, shifted gradually and measured honestly. Nobody sensible argues for one and not the other. The argument is really about sequencing, and about which side is currently paying for the other’s results. Performance marketing harvests demand. Brand marketing creates it. Run either alone for long enough and the other’s absence shows up in your numbers, usually as a cost you blame on the platform. ## What each one is actually doing A search ad against a category keyword is not creating a buyer. It is intercepting one, and the price of that interception is set by everyone else trying to do the same thing. When your brand is well known, you intercept more cheaply, convert better and get chosen at parity. That discount is what brand work buys, and it is invisible in the channel report, which is why it keeps getting cut. Conversely, brand investment with no capture layer produces demand you then hand to a competitor with a better landing page. Awareness that nobody can act on is a gift to the market. ## The diagnostic You are short of brand if: cost per acquisition rises every year on flat volume, your branded search is flat while category search grows, you win on price more often than on preference, and your sales team keeps asking for discounts. You are short of performance if: branded search is rising but revenue is not, people describe you accurately and still buy elsewhere, and your site converts badly on traffic that arrived with intent. Both diagnoses are answerable from data you already have. Share of search (https://colleviate.com/blog/share-of-search-leading-indicator.html) against your competitive set is the cheapest single indicator; it moves before revenue does, in both directions. ## About the 60:40 number The widely cited long-run split — roughly 60 percent to brand, 40 percent to activation — comes from effectiveness research on established consumer brands. It is a useful anchor and a bad rule. A two-year-old D2C brand with a working acquisition engine is right to invert it; a twenty-year-old category leader spending 90 percent on retargeting its own customers is not. The direction of travel matters more than the ratio: as you grow, the share that has to go to creation rises, because you eventually exhaust the demand the category generates for you. ## How to shift the ratio without losing the quarter - Fund the shift from efficiency, not from volume. There is almost always 10 to 20 percent of capture spend doing nothing except claiming credit for demand that would have converted anyway. Brand bidding at saturation and broad retargeting are the usual suspects. - Make the brand work do double duty. The same production can serve a fifteen-second brand cut and a six-second product cut. This is a scheduling decision, not a budget one. - Change the reporting before the budget. If the dashboard only rewards last click, every brand rupee will look like waste, and the shift will be reversed in month three. - Hold it for four quarters. Brand effects accumulate slowly and are indistinguishable from noise before that. Deciding in month two is how most attempts die. That is a governance problem, and we have written about it in audacity and tenacity (https://colleviate.com/blog/audacity-and-tenacity.html). ## The measurement trap Attribution models are built to allocate credit, not to find truth. Anything that creates demand rather than captures it will be under-credited by design, because it happens earlier, off-platform, and sometimes years before the purchase. That is an argument for using incrementality tests and holdouts, not for pretending the effect is not there. See what ROAS actually tells you (https://colleviate.com/blog/performance-marketing-roas-reality.html). ## The practical version Decide what share of next year buys demand you do not yet have. Write it down with the reasoning. Protect it from the quarterly instinct to move it into something that reports well. Then judge it on leading indicators for a year before judging it on revenue. That is the whole discipline, and it is mostly a test of nerve rather than of analysis. ## Where does the 60:40 number come from? From Les Binet and Peter Field's work for the IPA, published as The Long and the Short of It in 2013, which analysed campaigns entered for the IPA Effectiveness Awards over three decades. Across that databank, the most effective campaigns on average balanced roughly 60 percent of investment in long-term brand building with 40 percent in short-term activation. The authors themselves present it as an average across many categories and situations, to be adjusted rather than copied. B2B marketing tends to sit well away from it. The B2B Effectiveness Code, published in 2021 by LinkedIn's B2B Institute with WARC and Cannes Lions from 435 B2B case studies, found that B2B campaigns skew heavily towards short-term, rational and tightly targeted activity aimed at immediate sales effects, and that campaigns became more effective with more spend, longer duration and broader media. That supports the argument above: the ratio is an anchor, and the direction of travel matters more than the number. Related. What advertising does to awareness and what it does not do to belief: does advertising build awareness, or belief? (https://colleviate.com/blog/does-advertising-build-awareness-or-belief.html). ## Sources - The next chapter for The Long and the Short of It, IPA, undated: https://ipa.co.uk/knowledge/ipa-blog/the-next-chapter-for-the-long-and-the-short-of-it/ - The B2B Effectiveness Code, LinkedIn B2B Institute, WARC and Cannes Lions, 2021: https://business.linkedin.com/marketing-solutions/b2b-institute/the-b2b-effectiveness-code ## Questions **Does brand marketing work for B2B?** Yes, and the mechanism is the same: being known and credible before the buying cycle starts lowers the cost of every later step. The difference is the buying group is smaller and the cycle is longer, so effects take longer to read. **How do you measure brand investment?** Leading indicators first: share of search, unprompted recall, direct traffic, and the trend in branded query volume. Then incrementality tests such as regional holdouts, which are more honest than any attribution model. **Can a small company afford brand work?** Brand work is not a budget size, it is a decision about consistency. A small company with one distinctive idea held for three years is doing brand work. One that relaunches annually is not, however much it spends. --- # Marketing a clinic or diagnostic chain URL: https://colleviate.com/blog/marketing-for-clinics-and-diagnostics.html Topic: Growth audits | Published: 2026-09-14 | Updated: 2026-09-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Patient acquisition for a clinic is almost entirely a local search and response-time problem. Complete the profiles, publish procedure and price information, make booking take under a minute, and keep health interest data out of advertising platforms. Somebody searching at eleven at night for a diagnostic near them is not comparing brands. They are comparing which one looks open, close and competent. Clinics compete within a radius. That single fact decides most of the marketing plan: visibility in local search, a profile that answers the practical questions, and a booking route that works on a phone in thirty seconds. ## The profile is the shopfront Hours that are accurate including holidays, services listed individually, departments and specialisations, parking, real photographs of the actual premises, and current reviews. Most clinic enquiries begin and often end here, before anybody reaches the website. ## Publish procedures and prices Patients want to know what a test costs, what the procedure involves, how long results take and whether their insurance applies. Clinics that publish this rank for exactly those searches and convert better, because the uncertainty is what delays the booking. The pages also answer the questions reception spends its day repeating. ## Booking friction Every extra field, login or callback promise loses patients to the clinic that allowed a direct slot selection. Where a call is still the main route, answer it: an unanswered phone at a clinic is a lost patient, not a missed call. ## Reviews, handled carefully Ask every satisfied patient, reply to all of them, and never confirm or deny that a named individual was treated, even when their own review says so. Move specifics to a private channel and say publicly that you have. The wider reputation mechanics are in reputation is a search result (https://colleviate.com/blog/reputation-is-a-search-result.html). ## Data, strictly Health interest is sensitive. Do not pass condition names, procedure pages or appointment data into advertising platforms through URLs, event names or audiences, keep analytics separate from advertising tags, and be able to show what you collect and why. The implementation detail is in healthcare marketing without overclaiming (https://colleviate.com/blog/healthcare-marketing-india.html). ## Questions **What is the most effective channel for clinics?** Local search, by a distance. A complete profile with accurate hours, services, photographs and recent reviews outperforms most paid campaigns in the same radius. **Should clinics publish prices?** Yes where you can. Cost and process uncertainty is what delays bookings, and the pages rank for exactly the searches patients make. --- # What ROAS actually tells you, and what it hides URL: https://colleviate.com/blog/performance-marketing-roas-reality.html Topic: Measurement | Published: 2026-09-13 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** ROAS is attributed revenue divided by campaign cost, and every term flatters it: attribution is a modelling choice, revenue is usually gross rather than contribution, and cost excludes creative, agency fees and funded discounts. Report blended marketing efficiency ratio, total revenue over total marketing spend, beside platform ROAS, and test incrementality. A 6x return reported inside an ad platform is a claim made by the party being paid. It is not a lie, but it is not evidence of profit either. Return on ad spend is revenue attributed to a campaign divided by what that campaign cost. Every part of that sentence is doing more work than it appears to. “Attributed” is a modelling decision. “Revenue” is usually gross, not contribution. And “cost” typically excludes the creative, the agency and the discount you funded to close the sale. ## Three reasons the platform number is high - Attribution windows. A seven-day click and one-day view window will credit purchases that a stricter window would not. Compare two campaigns on different windows and you are comparing nothing. - Double counting. Google and Meta will both claim the same conversion. Add up platform-reported revenue across channels and it routinely exceeds what the finance system recorded. - Demand you already had. Retargeting and brand-term bidding intercept people who were coming anyway. Their reported return is excellent and their incremental return is often close to zero. ## The number to put beside it Marketing efficiency ratio — total revenue divided by total marketing spend — has one great virtue: it cannot be gamed by attribution, because neither term is attributed. It is blunt, it says nothing about which channel worked, and it is the honest headline. Track blended MER against platform ROAS. When they diverge, the platform is telling you a story about credit, not about growth. Then add contribution margin. A 4x return on a product with 25 percent margin loses money. Plenty of D2C brands scaled for two years on a ROAS target that was structurally below break-even, which is an expensive way to learn what a gross margin is. ## Incrementality, cheaply The only reliable way to know whether spend caused revenue is to withhold it and watch. You do not need a research budget for this: - Geographic holdouts. Turn a channel off in two comparable cities for four weeks. Compare against matched cities. Crude, but it answers the actual question. - Brand-term test. Pause bidding on your own name for a fortnight. Most brands find organic absorbs the majority of that traffic. Some find it does not, which is also worth knowing. - Spend-step tests. Move budget up or down by 30 percent and watch total orders, not attributed ones. Small changes teach you nothing; noise swallows them. ## Why targets drift A ROAS target set once and never revisited becomes a ceiling on growth. Efficiency and volume trade against each other: the efficient spend is by definition the spend against people closest to buying, and there are only so many of them. Insisting on a fixed ROAS while demanding growth is asking a team to find more of the cheapest demand than exists. The better frame is a contribution target at a given volume, revisited quarterly. That lets you buy expensive customers deliberately when the payback justifies it, instead of pretending they are cheap ones. ## What to report to a board Blended MER and contribution, incremental results from any test you ran, cost per acquisition by cohort rather than by campaign, and payback period. Platform ROAS belongs in the channel review, not in the board pack, for the same reason you would not take a vendor’s own estimate of their value to the audit committee. None of this makes paid media less valuable. It makes it legible, which is a precondition for spending more on it with confidence. If your tracking cannot support this, fix that first: see a GA4 setup that survives scrutiny (https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html) and the measurement gap (https://colleviate.com/blog/marketing-measurement-gap.html). ## How far off can platform attribution be? Far enough to change decisions. In a study published in Marketing Science in 2019, Gordon and colleagues compared fifteen randomized advertising experiments at Facebook with the observational methods advertisers commonly use, and found the observational estimates frequently failed to match the experimental ones, even with extensive controls. ROAS reported by a platform is an observational estimate made by the party being paid, which is why it belongs in its own column next to a measured number, not in the headline. ## Is there a cheap way to check? Yes, and it is getting cheaper. A geographic holdout, running a campaign in comparable markets and not in others, needs no platform cooperation and answers the only question that matters: what happened that would not have happened anyway. For a portfolio view, open-source mix models now cost nothing to run; Google made its Meridian model available to all marketers in January 2025. Neither replaces judgement, but both replace taking the platform's word for it. ## A worked example An illustrative account, to show the arithmetic rather than describe a market. A platform reports a ROAS of 6: Rs 60 lakh of attributed revenue on Rs 10 lakh of spend. Now apply the corrections this article describes. Contribution after product cost, fulfilment, returns and discounts is 35 percent of revenue, so the attributed sales contributed Rs 21 lakh. A holdout shows that half of those sales would have happened anyway, so the incremental contribution is Rs 10.5 lakh, against Rs 10 lakh of spend plus Rs 1.5 lakh of creative and fees. The campaign the dashboard called a 6x success lost money. None of the inputs were exotic. The difference came from three questions the platform report never asks: what did each sale actually contribute, what would have happened without the spend, and what did the campaign really cost. Related. The clearest published test of this problem is in what eBay learned when it switched off its brand ads (https://colleviate.com/blog/what-ebay-learned-by-switching-off-brand-ads.html). ## Sources - A comparison of approaches to advertising measurement: evidence from big field experiments at Facebook, Marketing Science 38(2), 2019: https://pubsonline.informs.org/doi/10.1287/mksc.2018.1135 - Google opens Meridian, its marketing measurement tool, to all users, Search Engine Land, January 2025: https://searchengineland.com/google-meridian-all-users-451286 ## Questions **What is a good ROAS?** There is no universal figure. Break-even ROAS is the inverse of your contribution margin, so a 40 percent margin business breaks even at 2.5x. Anything quoted without a margin attached is decoration. **Why do platform numbers exceed our actual revenue?** Because each platform claims conversions it touched, and their windows and models overlap. Reconcile against the finance system and use blended MER as the headline. **How often should we run incrementality tests?** Once or twice a year per major channel, and whenever you are about to make a large budget decision based on reported returns. --- # How to write a brief an agency can use URL: https://colleviate.com/blog/writing-a-brief.html Topic: Method | Published: 2026-09-13 | Updated: 2026-09-13 Publisher: Colleviate (https://colleviate.com/) **Short answer.** One objective, one audience, one thing to say, the real constraints, and the names of everyone who can say no. A brief with three objectives is not a brief, it is a negotiation that will happen later, at the presentation, in front of people. Most weak work traces back to a brief that was agreed by everyone because it asked for everything. A brief is a decision document. Its job is to remove options before anybody starts, which is why the painful version written in an afternoon saves a month. ## One objective Awareness, consideration, acquisition, retention, or a specific commercial number. Pick one. Three objectives produce work that is a compromise between them, and the compromise is always the safest option on the table. ## A real audience Not "men and women 25 to 45 in metros", which describes forty million people and excludes nobody. Describe someone specific enough that the creative team could picture them answering the phone. Include what they currently believe, since the job is to change that. ## One thing to say If the work has to communicate five points, it communicates none. Write the single sentence you want repeated back to you, and list the other four as supporting material that can appear where there is room. ## The real constraints Budget, as a number. Timings, with the immovable date named. Legal and regulatory requirements. Mandatory elements, honestly listed, including the ones somebody senior will insist on at the end whether or not they are in the brief now. ## Who can say no The most useful section and the one almost always missing. Name every person whose approval is required, in order. Work dies in the gap between the people who commissioned it and the people who turn out to have a veto, which is the subject of approvals that do not kill the work (https://colleviate.com/blog/approvals-that-do-not-kill-work.html). ## What does not belong The solution. If the brief describes the execution, you have hired an expensive production service and will get exactly what you asked for, which is rarely what you needed. The distinction is in the brief is not the idea (https://colleviate.com/blog/the-brief-is-not-the-idea.html). ## Questions **How long should a creative brief be?** One page plus appendices. Anything longer is usually unresolved disagreement written down rather than decisions taken. **What is the most commonly missing part of a brief?** The list of people who can say no. Work dies in the gap between whoever commissioned it and whoever turns out to hold a veto. --- # What a growth audit should contain, and what most leave out URL: https://colleviate.com/blog/growth-audit-what-it-should-contain.html Topic: Growth audits | Published: 2026-09-12 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A growth audit should find where growth is actually constrained, with evidence a sceptical CFO could defend. Start from the outside rather than your analytics, define the peer cluster, cover all seven dimensions, tag every number with its evidence tier, size the measurement gap, rank the plan by expected movement, and state how the audit could be wrong. An audit is the only part of an agency engagement where the incentives are clean. Nothing has been sold yet. So it is the part worth doing properly, and the part most agencies do worst. Ask five agencies for a growth audit and you will get five sales decks. Each will have a slide of screenshots, a spreadsheet of keywords, a few red boxes drawn around your homepage, and a recommendation that happens to match the service the agency sells. That is not an audit. It is a proposal wearing an audit's clothes. A growth audit has one job: to find where growth is actually constrained, with enough evidence that you could hand the document to a sceptical CFO and defend every line. Below is what that requires, in the order we do it. ## 1. Start from the outside, not from your analytics Your analytics tell you what happened to people who already found you. They say nothing about the people who found a competitor instead. So the audit has to begin where the market begins: search results, AI answers, review sites, app stores, job boards, pricing pages, and your competitors' equivalents of all of those. Reading a brand outside-in (https://colleviate.com/blog/outside-in-brand-audit.html) is the discipline; it means every finding is one a prospect could reproduce. ## 2. Define the peer cluster before you score anything A number without a comparison is an anecdote. A 2.1-second load time is fine in one category and a liability in another. Every metric in the audit should be scored against a named set of structural peers, chosen by business model and buyer, not by whoever the founder considers a rival. We resolve that cluster from public evidence, not from a list the founder writes down, so the comparison set is defensible. ## 3. Cover all seven dimensions, not just the one you sell This is where most audits fail. An SEO agency audits search. A paid agency audits ad accounts. A growth constraint does not respect service lines. A real audit covers market position, operational health, strategic posture, audience intelligence, measurement integrity, business and financial health, and whether the brand learns from any of it. A brand with excellent search visibility and a broken checkout is not a search problem, and an audit that only looks at search will spend your money in the wrong place. ## 4. Tag every number with its evidence tier Measured, real-user, sampled, modeled, inferred, observed, lab. A lab page-speed score and a real-user field score are different things and should never sit in the same sentence unlabelled. If a figure is modeled, the audit says so. Nothing erodes trust faster than a confident number that turns out to be an estimate. ## 5. Size the measurement gap Before recommending spend, the audit should say how much of your current spend is unmeasured or misattributed. In our experience it is rarely under a fifth of the budget. The measurement gap (https://colleviate.com/blog/marketing-measurement-gap.html) is usually the cheapest thing to fix and the most valuable, because it makes everything after it provable. ## 6. Rank the plan by expected movement, with the evidence attached The output is not a list of recommendations. It is a ranked 90-day plan where each line says what it should move, by roughly how much, against which baseline, and what evidence supports the expectation. Some lines will be uncertain, and the plan should say that too. ## 7. State how the audit could be wrong Every audit rests on assumptions about the buyer, the category, and the competitive set. The good ones say what would change the conclusion. That is the section a board actually reads. ## What it should not contain - A recommendation that matches only what the agency sells. If the constraint is your pricing page, an honest audit says so even when there is no retainer in it. - Unlabelled estimates. Every "industry benchmark" needs a source. - Screenshots as evidence. A screenshot shows that something exists. It does not show that it matters. - A timeline that starts on signature. The audit should already be dated, baselined, and re-runnable. An audit built this way takes about two weeks and is worth keeping even if you never hire the agency that wrote it. That is the test: would you pay for the audit on its own? If the answer is no, you are reading a proposal. Sector reading. The audit changes by sector: fintech (https://colleviate.com/blog/fintech-marketing-india.html), education (https://colleviate.com/blog/education-marketing-india.html), SaaS (https://colleviate.com/blog/saas-marketing-india.html), travel and tourism (https://colleviate.com/blog/travel-and-tourism-marketing.html) and wellness and fitness (https://colleviate.com/blog/wellness-and-fitness-marketing.html). ## Questions **How long should a growth audit take?** Two weeks is realistic for a single brand and its peer cluster when the outside-in collection is automated. Longer usually means manual work or a sales process, not more rigour. **Do you need access to our analytics for a growth audit?** No. The first pass is entirely outside-in, which is the point: it shows you what the market sees. Internal data is added afterwards to size the measurement gap and validate the baseline. **Is a growth audit the same as an SEO audit?** No. An SEO audit covers one channel. A growth audit covers every dimension where growth can be constrained, including measurement, positioning, pricing and operations, and ranks them against each other. --- # Data-driven marketing is not a dashboard URL: https://colleviate.com/blog/data-driven-marketing-definition.html Topic: Method | Published: 2026-09-12 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Marketing is data-driven when each decision was ranked against the alternatives by evidence, baselined before the money was spent, and re-scored afterwards with the evidence attached. A dashboard is reporting, a description of the past, not a decision. Plans assembled first and justified with numbers later do not qualify. The phrase has been used so widely that it no longer distinguishes anyone. Here is the definition we hold ourselves to, and the three tests that separate it from reporting. Reporting is a description of the past. A dashboard is reporting with colours. Neither one is a decision. Data-driven marketing, if the words mean anything, has to describe how decisions get made, not how results get displayed afterwards. Our working definition is short. A marketing decision is data-driven when it was ranked against the alternatives by evidence, baselined before the money was spent, and re-scored after, with the evidence attached to the result. Every clause does work. ## Test one: was the decision ranked, or just justified? Most marketing plans are assembled and then supported with data. The channel mix was decided in a meeting; the numbers were found later. A ranked decision starts the other way round: here are the twelve things we could do, here is what the evidence says each one would move, here is the order. The plan is whatever comes out of the ranking, including when it is inconvenient for the agency. ## Test two: was there a baseline before the spend? You cannot know what marketing changed if you do not know where the brand stood the day before. A baseline is not last month's dashboard. It is a scored, dated position across every dimension the plan might move, against the peers the brand is actually competing with. This is why every Colleviate engagement begins with an audit (https://colleviate.com/blog/growth-audit-what-it-should-contain.html) and not with a campaign: the audit is the baseline. ## Test three: was the result re-scored with evidence, or reported with a narrative? "Brand awareness improved" is a narrative. "Share of search moved from 23 to 27 percent against the same four peers, measured the same way as the baseline, while paid spend was flat" is a re-score. The second one can be wrong, which is exactly what makes it useful. A claim that cannot be wrong is not evidence of anything. ## What this rules out - Vanity metrics without a peer comparison. Impressions, followers and sessions are inputs, not outcomes, and they mean nothing without a competitive baseline. - Attribution by assertion. If a channel is credited with revenue, the model that credits it should be inspectable. Attribution (https://colleviate.com/blog/marketing-attribution-for-cmos.html) is a measurement problem before it is a marketing one. - Benchmarks without sources. "Industry average CTR" is an anecdote unless the industry, the period and the sample are stated. - Plans that cannot fail. Every line in a data-driven plan says what would show it did not work. ## Why most agencies stop at the dashboard Because a dashboard is cheap to build and impossible to be wrong about. A ranked, baselined, re-scored plan is expensive to build and easy to be wrong about. The second one is the only one that lets a client learn anything, and the only one that lets an agency improve. We built our practice so that the expensive version is affordable: the outside-in read, the peer cluster and the re-score are automated, which leaves the humans to do the ranking and the work. ## A simple way to check your current agency Ask for last quarter's plan and last quarter's results side by side. If the plan says what each line was expected to move and the results say what it actually moved, against the same baseline, you have a data-driven partner. If the plan is a list of activities and the results are a dashboard, you have a reporting partner. Both can be fine. Only one can be held to account. ## Questions **What is the difference between data-driven and data-informed marketing?** Data-informed usually means the data was consulted. Data-driven, as we use it, means the decision was ranked by evidence and can be checked against a baseline afterwards. The difference is whether the data could have changed the decision. **Do small brands have enough data for data-driven marketing?** Yes, because the first-party data is not the constraint. Outside-in signals such as search, reviews, pricing and competitor movement exist for every brand at any size, and they are what the baseline is built from. --- # SEO vs AEO: where search demand is going in 2026 URL: https://colleviate.com/blog/seo-vs-aeo-2026.html Topic: AI visibility | Published: 2026-09-12 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Search did not die in 2026; it split. Ranked links still carry most transactional intent, while synthesised AI answers carry much of the research before it. Almost all SEO foundations carry over to AEO, because answer engines retrieve from the same crawlable, fast, well-structured web. Run one programme measured on both surfaces. Search did not die. It split. Ranked links still carry most transactional intent; synthesised answers now carry most of the research that precedes it. A brand that only optimises for one surface is invisible for half the journey. The argument about whether AI answers are replacing search is mostly a distraction. Both surfaces exist, both are growing in different parts of the buying journey, and a brand needs to be findable on each. The useful question is narrower: what is different about earning visibility in an answer engine, and how much of the existing SEO programme carries over? ## What stays the same Almost all of the foundations. Answer engines retrieve before they generate, and what they retrieve is largely the same crawlable, fast, well-structured, authoritative web that ranking algorithms reward. Technical health, clear information architecture, genuine expertise and consistent entity signals are prerequisites for both surfaces. A site that ranks poorly for structural reasons will also be cited poorly. How answer engines choose sources (https://colleviate.com/blog/how-ai-answer-engines-choose-sources.html) covers the retrieval mechanics in detail. ## What changes | Dimension | Ranked search (SEO) | Answer engines (AEO) | | Unit of competition | The page, for a query | The passage, for a question | | What wins | Relevance plus authority plus experience | A quotable, specific, verifiable claim | | Outcome | A click | A citation, often without a click | | Measurement | Rankings, clicks, share of voice | Citation share (https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html) across engines, by question | | Failure mode | Ranking for the wrong intent | Being paraphrased without attribution | The practical consequence is that AEO rewards a writing style most marketing content avoids: direct claims, stated numbers, named comparisons, and definitions that can be lifted whole. Content built to keep a reader scrolling performs worse in an answer engine than content built to be quoted. ## Where the demand is actually moving Research-stage questions, comparisons, definitions and "how do I choose" queries are shifting toward synthesised answers fastest. Transactional and navigational queries remain overwhelmingly link-based. For most B2B and considered-purchase brands, that means the top of the funnel is increasingly decided in an answer engine, while the bottom is still decided on a results page. Losing the first does not show up in analytics, because the visits that never happened leave no trace. It shows up in share of search (https://colleviate.com/blog/share-of-search-leading-indicator.html), which is why we track it as a leading indicator. ## One programme, two surfaces The efficient approach is not two teams. It is one content and technical programme with two measurement layers: - Baseline both. Rankings and clicks for the query set; citation share across ChatGPT, Gemini, Perplexity and Google AI Overviews for the question set. Do this against named peers, not in isolation. - Fix the shared foundations first. Crawlability, speed, structured data, entity consistency, and one authoritative page per topic. This lifts both surfaces. - Rewrite for quotability where the question set demands it. Lead with the answer, state the number, name the comparison. Keep the depth below it for the human reader. - Earn third-party corroboration. Answer engines weight claims that appear in more than one credible place. Coverage, reviews and directories matter more for AEO than they did for classic SEO. - Re-score monthly on both layers. Citation share moves faster than rankings and is more volatile. Read the trend, not the week. ## What to tell the board That search visibility is now two numbers, that the second one is not in the analytics, and that the brand has a baseline for both as of a stated date. The rest is execution. ## What happened to the 25 percent prediction? In 2024 Gartner predicted that by 2026 traditional search engine volume would drop 25 percent, with search marketing losing share to AI chatbots and other virtual agents. It was a forecast, made in a press release, and we have not seen published data that confirms or refutes the exact figure. What has been measured points the same way. Alphabet reported in July 2025 that AI Overviews had more than 2 billion monthly users across more than 200 countries and territories. Pew Research Center found that when an AI summary appeared, people clicked a traditional result in 8 percent of visits, against 15 percent when it did not. So the useful framing for a board is not whether search volume fell by a quarter, but that more questions are now answered above the links, and fewer people click through. That is the case for measuring both surfaces rather than one. ## How do you measure both surfaces? With two instruments, read side by side every month. For the ranked surface, Search Console already gives you impressions, clicks and average position for the queries that matter. For the answer surface, build a fixed panel: the twenty to fifty questions your buyers actually ask, run through the main engines on a schedule, recording whether your brand is named, whether the description is accurate, and which competitor is named instead. Keep the questions fixed so the trend means something. Then add one leading indicator that sits above both: share of search (https://colleviate.com/blog/share-of-search-leading-indicator.html) for your competitive set, which tells you whether the demand both surfaces serve is moving towards you. Three numbers, one page, reviewed together. Reported separately, each one flatters a different team. Related. Buyers now ask assistants who to hire as well as searching; how B2B buyers shortlist agencies (https://colleviate.com/blog/how-b2b-buyers-shortlist-agencies.html) looks at when that happens in the buying process. ## Sources - Will search engine traffic really drop 25% by 2026, as Gartner predicts?, CMSWire, March 2024: https://www.cmswire.com/digital-experience/will-search-engine-traffic-really-drop-25-by-2026-as-gartner-predicts/ - Alphabet Q2 earnings call: CEO’s remarks, Google, July 2025: https://blog.google/company-news/inside-google/message-ceo/alphabet-earnings-q2-2025/ - Google users are less likely to click on links when an AI summary appears in the results, Pew Research Center, 22 July 2025: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/ ## Questions **Is AEO replacing SEO?** No. They cover different parts of the journey. Research and comparison questions are moving to answer engines; transactional and navigational searches remain link-based. A brand needs both. **How do you measure AEO?** Citation share: for a defined set of questions your buyers ask, how often each engine cites you versus named competitors, tracked over time. It has to be measured across several engines because they retrieve differently. **Does AEO require different content?** Partly. The foundations are shared. The difference is quotability: direct answers, stated numbers and named comparisons near the top of the page, with depth below. --- # Marketing attribution for CMOs who have to defend the budget URL: https://colleviate.com/blog/marketing-attribution-for-cmos.html Topic: Measurement | Published: 2026-09-12 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Every attribution model is wrong, so a CMO needs one that is wrong in a way the CFO can inspect. Choose the model last. Start from an outside-in baseline of public signals, such as share of search and AI-answer citation share, that tracking changes cannot break; then count only what is deterministically tracked, with the gaps stated, and report three numbers a board will accept. Every attribution model is wrong. The useful ones are wrong in a way you can inspect. If the CFO cannot see how a channel was credited, the credit will not survive the next budget round. The attribution conversation usually starts in the wrong place: which model. First-touch, last-touch, linear, data-driven, whichever the ad platform prefers this quarter. The model is the last decision, not the first. Before it come three questions a board will actually ask, and most marketing reporting cannot answer any of them. ## The three questions - How much of the spend is measured at all? Not modeled, not assumed. Measured. In our audits the answer is rarely above four-fifths, and the unmeasured fifth is where the arguments live. Start by sizing the measurement gap (https://colleviate.com/blog/marketing-measurement-gap.html) honestly. - What would have happened without the spend? This is incrementality, and it is the only attribution question a finance team truly cares about. Platform-reported conversions answer a different question: which conversions did the platform touch. Those are not the same number and the difference is usually large. - Can someone outside marketing reproduce the number? If the credit for a channel depends on a setting inside a vendor dashboard, the number belongs to the vendor. A board-grade attribution figure has a source, a method, and a confidence, and it can be recomputed by someone who did not build it. ## Rebuilding attribution on evidence The approach we use has four layers, and the order matters. ### Layer one: the outside-in baseline Before any internal data, score the brand's demand position from public signals: share of search against peers, AI-answer citation share (https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html), review velocity, direct-traffic proxies. These are independent of your tracking setup, which means they cannot be broken by a consent banner or a cookie change, and they move before revenue does. They are the control group your internal data lacks. ### Layer two: the measured floor Count only what is deterministically tracked, end to end, with the gaps stated. This number will be smaller than what the platforms report. That is the point. It is the floor you can defend without a caveat. ### Layer three: modeled lift, labelled as modeled Geo tests, holdouts, and marketing-mix models fill the space above the floor. Each figure carries its evidence tier: measured, modeled, or inferred. A board will accept a modeled number if it is labelled as one and the method is stated. It will not accept a modeled number presented as measured, and it will remember when it finds out. ### Layer four: the re-score Attribution is not a report, it is a loop. Every month the outside-in signals, the measured floor and the modeled lift are re-scored against the same baseline. Movement in the outside-in layer that is not explained by spend is the most interesting finding in the pack, because it points at something the plan missed. ## What to stop reporting - Platform-attributed revenue as a headline. Report it, but as the platform's claim, in its own column, next to the measured floor. - Blended ROAS without a spend split. One good channel hides three bad ones. - Any metric without last quarter's value beside it. A number without a baseline is decoration. ## The three numbers a board will accept Measured revenue with the gap stated. Modeled incremental lift with the method stated. Outside-in demand position against named peers, dated. Present those three, in that order, every quarter, and the attribution argument stops being about the model. It becomes a conversation about what to do next, which is the conversation a CMO actually wants to be in. Related. What counts as evidence in the first place: data-driven marketing, defined (https://colleviate.com/blog/data-driven-marketing-definition.html). ## Questions **Which attribution model is best?** None of them is best in the abstract. The question is whether the model is inspectable and whether it is presented alongside a measured floor and an outside-in baseline. A simple model labelled honestly beats a sophisticated one presented as fact. **What is the difference between attribution and incrementality?** Attribution assigns credit for conversions that happened. Incrementality asks how many of those conversions would have happened anyway. Finance teams care about the second one. --- # How to brief a performance marketing agency you can actually measure URL: https://colleviate.com/blog/how-to-brief-a-performance-marketing-agency.html Topic: Method | Published: 2026-09-12 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Brief a performance marketing agency as a measurement contract. Define the baseline across the dimensions the work should move, name the four to six real competitors, set what success and failure look like, and agree how results will be checked. If no baseline exists, commission it first as a separate, paid audit. Most agency relationships fail in the brief, months before anyone notices. If the brief does not define the baseline, the peer set and the failure condition, no report the agency writes afterwards can be checked. A brief is usually written as a description of the client and a list of things the client would like. It is rarely written as a measurement contract. That is why, a year in, the client cannot say whether the agency worked and the agency cannot prove that it did. Both sides are arguing about a number nobody defined. ## Five things to put in the brief ### 1. The baseline, dated Where the brand stands today, across the dimensions the work is meant to move, measured in a way that can be repeated. If you do not have this, make the first deliverable an audit (https://colleviate.com/blog/growth-audit-what-it-should-contain.html) that produces it, and make the audit a separate, paid piece of work so the agency has no incentive to shade it. ### 2. The peer set Name the four to six companies the brand actually competes with for the buyer's attention, and require that every metric be reported against them. Absolute numbers flatter everyone. Relative numbers tell the truth. ### 3. The constraint, not the channel Brief the problem: demand is flat in one segment, conversion drops at pricing, the brand is invisible in AI answers. Do not brief the channel. An agency briefed on "run paid social" will run paid social whether or not it is the constraint. An agency briefed on the constraint has to argue for its recommendation with evidence. ### 4. The evidence standard State that every reported figure must carry its source and its evidence tier: measured, modeled or inferred. State that platform-attributed conversions will be reported as the platform's claim, in their own column. This one paragraph changes the quality of every report you will receive. ### 5. The failure condition What result, by what date, would show that the approach is not working and should change. Write it down before the work starts. An agency that will not agree to a failure condition is telling you something. ## Three things to leave out - A required channel mix. That is the agency's job to recommend and yours to interrogate. - Vanity targets. Followers, impressions and sessions are inputs. Brief outcomes. - A twelve-month scope on day one. Brief the audit and the first ninety days. Renew on evidence. ## The one question At the pitch, ask: show me a client where your recommendation was to spend less, and what happened. A partner has an answer. A vendor changes the subject. The reason this question works is that it tests whether the agency's process can produce an answer the agency does not profit from. Everything else in the brief is there to make sure it keeps doing so once the contract is signed. ## What a good response looks like It pushes back on the brief. It asks for the baseline or proposes to build one. It names the peer set and may disagree with yours. It ranks the possible work by expected movement and says which lines are uncertain. It states the failure condition in its own words. And it is shorter than you expected, because most of what usually fills an agency response is there to avoid being pinned down. Related. For the brief itself, see how to write a brief (https://colleviate.com/blog/writing-a-brief.html). ## Questions **Should the audit be paid separately from the retainer?** Yes. A free audit is a sales tool and will recommend the service on sale. A paid audit can recommend that you spend less, or elsewhere, which is the only kind worth having. **How long should the first agency engagement be?** Audit plus ninety days. That is long enough to move outside-in signals and short enough that a wrong approach is cheap to change. --- # Approvals that do not kill the work URL: https://colleviate.com/blog/approvals-that-do-not-kill-work.html Topic: Method | Published: 2026-09-12 | Updated: 2026-09-12 Publisher: Colleviate (https://colleviate.com/) **Short answer.** One decider, named in advance. Everybody else advises. Collect all feedback in a single round against the brief rather than against taste, and make anybody asking for a change say which objective it serves. Nothing good has ever been improved by its eleventh reviewer, and every organisation knows this while maintaining eleven reviewers. Approval processes are where brave work becomes safe work, not through a decision anybody made but through accumulation. Each individual comment is reasonable. The sum is a campaign nobody will notice. ## One decider Name the person who decides before the work is made. Everyone else contributes a view, which the decider may take or leave. Consensus review guarantees the lowest common denominator, because the only version everybody can agree on is the version that risks nothing. ## Feedback against the brief, not against taste "I do not like the blue" is a preference. "This does not say the one thing we agreed to say" is feedback. Require every request to name the objective it serves, and a large share of comments disappear before they are sent, which is the point. ## One round, everybody together Serial review is the real killer: creative goes to one person, is amended, goes to the next, who objects to the amendment. Get everyone in one room or one document at the same time, resolve contradictions there, and send a single consolidated response. ## Protect the thing that makes it work Most work has one element doing the heavy lifting: an uncomfortable line, an unusual visual, a joke. It will attract the most comments, because it is the most noticeable. Decide at briefing which parts are permanent, so the discussion in week six is about whether evidence has changed rather than whether somebody has got used to it. That argument is in audacity and tenacity (https://colleviate.com/blog/audacity-and-tenacity.html). ## Late objections Somebody senior who was not in the process will object at the end. The fix is structural: get them into the brief approval, where their input is cheap, rather than the final review, where it costs a reshoot. ## Questions **How many people should approve creative work?** One decider, with as many advisers as useful. Consensus approval reliably produces the safest version on the table. **How do you handle a senior objection at the final stage?** Prevent it by getting that person into brief approval. At final stage the only options are expensive changes or an uncomfortable conversation. --- # SEO services: what to expect, and what to refuse URL: https://colleviate.com/blog/seo-services-what-to-expect.html Topic: Method | Published: 2026-09-11 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A real SEO service covers three things: making the site legible to crawlers, making content worth ranking, and earning references from others. Technical fixes can show within weeks; new content on an established domain usually takes eight to sixteen weeks; a new domain takes six to twelve months. Refuse anyone promising page one in ninety days. Search is the only channel where the asset keeps working after you stop paying. It is also the one where the gap between competent and cosmetic work is hardest for a buyer to see. An SEO engagement has three parts, and most disappointing ones are missing the same part. The three are: making the site legible to crawlers, making the content worth ranking, and making other people reference it. Fixing only the first produces a technically immaculate site nobody links to. ## What a real scope contains - Technical foundation. Crawlability, index coverage, site speed, structured data, internal linking (https://colleviate.com/blog/internal-linking-for-topic-clusters.html), canonical hygiene, sensible URLs. Mostly finite work with a real end. - Content that answers a query better than what ranks now. Not more words. Better coverage of the question, by someone who knows the subject. - Internal linking as an ongoing habit. The cheapest ranking lever most sites ignore: every new page should link to and from the pages it belongs with. - Digital PR or genuine citation. Links from places with editorial standards. Slow, expensive, and the part that actually separates competitors. - Measurement. Rankings by cluster, non-branded organic sessions, assisted pipeline. Reported monthly, judged quarterly. ## Timelines, honestly Technical fixes can show within weeks on an established domain. New content on an established domain typically takes eight to sixteen weeks to find its position. A new domain takes longer than anyone wants to hear: six to twelve months before organic is a meaningful channel, and that assumes real content and real links. Any proposal promising page one in ninety days on a new site is selling either luck or something you will have to clean up. ## What to refuse - Link packages with a monthly quantity. Links you can buy by the dozen are links search engines can identify by the dozen. - Content priced by word count. Paying per thousand words guarantees you get thousands of words. - Keyword stuffing dressed as optimisation. Exact-match phrases in headings no human would write. - Reporting on ranking alone. Ranking for terms nobody searches, or terms that never convert, is an activity metric. - Anything described as a secret. There are no secrets, only work people have not done. ## The keyword question High-volume head terms are the ones clients ask for and the ones least likely to pay back. A term like “digital marketing” carries enormous volume and almost no commercial intent; the people searching it are mostly students and competitors. The money sits in the specific, lower-volume, high-intent phrases: the service plus the city, the problem plus the industry, the comparison, the price question. Build clusters around those and let the head term arrive as a by-product of depth. ## Where search is going A growing share of searches now end in a generated answer rather than a click. That does not remove the work, it changes what the work is for: being the source an answer engine cites rather than the tenth blue link. The disciplines overlap almost entirely — clear claims, verifiable sources, clean structure — which we set out in AEO and GEO (https://colleviate.com/blog/aeo-geo-ai-visibility.html). ## How to judge progress at month three Do not look at traffic. Look at whether the technical debt list is shrinking, whether new pages are indexed and picking up impressions, whether non-branded impressions are trending up in Search Console, and whether anything you published has been cited anywhere you did not pay for. Those four move before traffic does, and if none of them is moving by month three, traffic will not move by month six. ## Questions **How long does SEO take to work?** On an established domain, eight to sixteen weeks for new content to settle. On a new domain, six to twelve months before organic search is a meaningful channel. **Is link building still necessary?** Yes, but the unit is credibility, not quantity. A handful of references from publications with editorial standards outperforms hundreds of directory placements. **Should we target high-volume keywords?** Usually not first. Head terms carry volume and weak intent. Specific phrases convert better and are winnable, and depth across them tends to lift the head term anyway. --- # How to structure a small marketing team URL: https://colleviate.com/blog/marketing-team-structure.html Topic: Method | Published: 2026-09-11 | Updated: 2026-09-11 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Hire a commercially literate generalist who can own direction first, a content or creative owner second, and a performance or analytics owner third. Specialise only when a discipline has enough continuous work to fill a role, and buy it until then. The first hire decides the next five, because whatever they are good at becomes what the company thinks marketing is. Small teams fail in a predictable way: they hire for the task that is loudest this quarter, repeat that three times, and end up with three executors and nobody deciding what to execute. ## The first hire Senior enough to own direction and argue with sales and finance. Commercially literate, comfortable with numbers, able to write. Not a specialist in the channel that happens to be urgent, because urgency changes and the hire does not. ## The second Usually a content and creative owner, because that is the work with the most continuous volume and the highest cost to outsource per unit. In product-led or ecommerce businesses the second hire is often performance instead. Pick based on where the continuous work is, not where the excitement is. ## The third Analytics and operations: the person who owns measurement, the stack and the reporting that everyone else relies on. Teams usually hire this one last and wish they had hired it second. ## Buy the rest Design, film, PR, specialist channel work and anything seasonal. These need range or capacity rather than continuity, which is exactly the trade examined in agency or in-house (https://colleviate.com/blog/agency-or-in-house.html). ## The structures that stop working - Marketing as a service desk. When every department can submit requests directly, strategy disappears under tickets within two quarters. - Channel silos too early. Three people each owning one platform, with nobody owning the customer. - A head of marketing with no budget authority, which is an execution role with a misleading title. ## The signal to restructure When the team spends more time coordinating than making. That usually arrives around the sixth or seventh person and is a sign to split by customer journey stage rather than by channel. ## Questions **What should the first marketing hire be?** A commercially literate generalist senior enough to own direction. Hiring an executor first leaves nobody deciding what to execute. **When should a team hire a specialist?** When that discipline has enough continuous work to fill a role. Below that, buying the capability is cheaper and usually better. --- # Google Ads or Meta Ads: where should the first rupee go? URL: https://colleviate.com/blog/google-ads-vs-meta-ads.html Topic: Measurement | Published: 2026-09-10 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** If people already search for what you sell, start with Google Ads: existing demand is cheaper to serve than new demand is to create, and it shows within weeks whether the offer converts. If nobody searches for your category yet, start with Meta, where social advertising creates the demand that search can later harvest. Check search volume first. The two platforms are not competitors in the way the question implies. One finds people already looking for you. The other finds people who would want you if they knew. Search advertising is interception. Someone has already framed their need as a query, and you pay to be in front of it. Social advertising is interruption, in the neutral sense: you choose the audience, and the creative has to earn the attention it takes. ## The decision rule If people already search for what you sell, start with search. Existing demand is cheaper to serve than new demand is to create, and it tells you within weeks whether the offer converts. If nobody searches for your category — a new product, a novel service, a problem people have not named — search has nothing to harvest, and social is where you create the demand you will later harvest. Check this before deciding, not after. Search volume for your category terms is free to look up, and a category with almost no query volume is telling you something important about where the first budget belongs. ## What each one is good at Search: high intent, measurable, limited by the size of existing demand, and steadily more expensive as competitors bid on the same intent. Excellent for services with a clear name, comparison queries, and anything urgent. Meta: enormous reach, cheap impressions, strong for visual products and impulse categories, and dependent on creative quality to a degree that surprises people arriving from search. It also creates demand that later shows up as branded search, which is why the two channels flatter and starve each other in reporting. ## What goes wrong on each - On search: bidding broadly and letting the algorithm find irrelevant queries; sending all traffic to the homepage; paying for your own brand name at saturation; judging by clicks rather than qualified enquiries. - On social: treating creative as an afterthought; running one asset until it exhausts the audience; optimising for cheap leads nobody can close; fixing an audience problem with a budget increase. On search in particular, the landing page does most of the work the ad gets blamed for. A strong ad against a weak page is an expensive way to test copy: see landing pages that convert (https://colleviate.com/blog/landing-pages-that-convert.html). ## Running both Most businesses past their first year should. The useful structure is: search covers the demand that exists, social creates the demand that does not, and retargeting stays deliberately small because it mostly re-bills people who were already coming. Measure the pair together with blended efficiency rather than comparing their self-reported returns, for the reasons in what ROAS actually tells you (https://colleviate.com/blog/performance-marketing-roas-reality.html). ## Budget floors Below about Rs 1 lakh a month, split across both platforms and you will learn nothing on either: neither gets enough data to optimise and neither gets enough creative to be fair to. Pick one, run it properly for a quarter, then add the other. At small budgets, learning rate matters more than coverage. ## The uncomfortable conclusion In most accounts we look at, the largest available gain is not in the platform choice. It is in the creative on social and the landing page on search, the two things clients most often treat as already solved. Channel arguments are easier to have than creative ones, which is why they take up more of the meeting. Related. The mechanics are in how to structure a Google Ads account (https://colleviate.com/blog/google-ads-account-structure.html), how to test Meta ad creative (https://colleviate.com/blog/meta-ads-creative-testing.html) and where retargeting stops working (https://colleviate.com/blog/retargeting-limits.html). ## Questions **Which is cheaper, Google or Meta?** Meta almost always has the lower cost per click and the higher cost per qualified lead. Search costs more per click because the intent arrives with it. Compare cost per qualified outcome, not per click. **Can we run both on a small budget?** Below roughly Rs 1 lakh a month, splitting starves both. Run one properly for a quarter and add the second once the first is stable. **How much creative does paid social need?** More than most plans allow. Expect to refresh the main assets every four to six weeks on an active account, and to test several distinct ideas rather than variations of one. --- # A marketing stack a small team can actually run URL: https://colleviate.com/blog/martech-stack-for-small-teams.html Topic: Method | Published: 2026-09-10 | Updated: 2026-09-10 Publisher: Colleviate (https://colleviate.com/) **Short answer.** You need somewhere to keep contacts, somewhere to send messages, a way to measure, and a place to plan work. Four tools, chosen for integration with each other rather than for features, and nothing else until a specific job is being done badly by a person. Most small-team stacks are a graveyard of annual subscriptions bought to solve problems that were never technical. Tools do not create process. They encode whatever process already exists, which is why a team with an unclear way of working buys software and ends up with an unclear way of working plus a renewal date. ## The four things you actually need - A contact database that both sales and marketing use. One, not two. - A messaging tool for email and, in India, usually WhatsApp. - Measurement: analytics, tag management, and the ability to reconcile against the finance system. - Work planning, which can be whatever the company already uses rather than a marketing-specific tool nobody else opens. ## Choose for integration, not features The feature comparison is almost never the thing that bites. What bites is that the email tool and the CRM disagree about who a contact is, and somebody spends Friday afternoons reconciling them. Pick tools that already talk to each other properly, and accept a weaker feature set for it. ## What to delay Attribution platforms before you have enough volume for the model to mean anything. Customer data platforms before you have more than two data sources. Personalisation engines before you have content worth personalising. Each of these solves a real problem that most small teams do not yet have. ## The cost nobody budgets Implementation and maintenance time. A tool that saves four hours a month and takes six hours a month to keep current is a net loss that will never appear in any review, because nobody measures it. ## An annual clear-out List every subscription, who uses it, and what would break if it were cancelled. In most teams at least a quarter of the stack fails that test. The money is better spent on the measurement layer, which is the one most under-resourced: see a GA4 setup that survives scrutiny (https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html). ## Questions **What marketing tools does a small team need?** A single contact database, a messaging tool, a measurement layer and somewhere to plan work. Everything else should wait for a specific, demonstrated need. **How do you choose between similar marketing tools?** On how cleanly they integrate with what you already run. Feature gaps are survivable; data that disagrees between systems is not. --- # Local SEO for service businesses: what moves the map URL: https://colleviate.com/blog/local-seo-for-service-businesses.html Topic: Method | Published: 2026-09-09 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Local SEO for a service business depends on relevance, distance and prominence. Distance you cannot change; the rest you can, by completing every field of the Google Business Profile, keeping name, address and phone identical everywhere, asking for reviews continuously, and building one page per service. Bulk directory citations and keyword-stuffed business names are a waste or a risk. Local search rewards a short list of unglamorous things done properly. Most businesses do four of them and wonder why the fifth competitor keeps appearing above them. For a service business, the map results are often worth more than the rest of the page combined. They sit above the organic listings, they carry reviews, and the people using them are usually minutes from deciding. ## What the ranking depends on Three things, roughly in this order: relevance, distance and prominence. Distance is the one you cannot change, which is why a competitor two streets from the searcher will sometimes beat you regardless of effort. The other two are entirely in your hands. - Relevance. Your profile category, services and description should match the words people actually search. The primary category matters more than any other single field. - Prominence. Reviews, consistent listings across directories, links, and the strength of the website behind the profile. ## The list, in order - Claim and complete the Google Business Profile. Every field: correct primary category, accurate hours, service areas, services listed individually, photographs that are yours and recent. - Fix name, address and phone consistency everywhere. The same string wherever it appears. Old suite numbers and dead phone numbers on directory listings are a common, invisible drag. - Ask for reviews, continuously. Volume matters, recency matters more than people expect, and replies matter because they are public evidence that a human is home. Never filter requests by expected sentiment; it breaches platform policy and it is obvious. - Build one page per service and per location. Not a single page listing everything. A dedicated page can rank for the specific query; generic pages rank for nothing in a competitive city. - Add the right structured data. LocalBusiness markup with address, hours and coordinates, so the machine reading of your page matches the profile. ## What is mostly a waste of time Buying citations in bulk from low-quality directories. Posting daily updates nobody reads. Stuffing the business name with keywords, which is a policy violation and periodically gets listings suspended. And judging your ranking from your own office, which is measured from your location and says little about what anyone else sees. ## Competing in a dense city In Mumbai or Bengaluru, dozens of competitors sit within the same few kilometres, so proximity stops being a differentiator and prominence decides it. That means reviews at volume, genuine local coverage, and service pages with real depth. It also means being honest about service areas: one office cannot rank across an entire metro, and pretending otherwise produces pages that rank nowhere. ## How to measure it Profile insights for calls, direction requests and website clicks, tracked monthly. Rank tracking from a grid of points rather than one. And, most usefully, ask every enquiry where they found you: for local businesses, self-reported attribution is often more accurate than the analytics. Reviews also shape what people and assistants say about you elsewhere, which we cover in reputation is a search result (https://colleviate.com/blog/reputation-is-a-search-result.html). ## Questions **How many reviews do we need to rank in the map results?** There is no threshold. What matters is being credible against the specific competitors in your radius: comparable volume, better recency, and replies that read like a person wrote them. **Does a service area business need a physical address?** Google requires a real address for verification, but you can hide it and show service areas instead. Faking a location is the fastest route to a suspended listing. **Can one website rank for several cities?** Yes, with a genuine page per city containing real local content. Duplicating one page with the city name swapped has not worked for years. --- # Scoping a retainer so it still works in month nine URL: https://colleviate.com/blog/scoping-a-retainer.html Topic: Method | Published: 2026-09-09 | Updated: 2026-09-09 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Scope by capacity and named deliverables together: how many people at what seniority for how many hours, and what that buys each month. Agree what happens when a request falls outside it before the request arrives, and review the whole scope quarterly. Retainers rarely fail at the start. They fail in the quarter where both sides quietly believe the other is taking advantage. A retainer buys continuity, which is genuinely valuable and almost impossible to specify precisely. The gap between what the client assumed and what the agency assumed is where the relationship goes. ## Scope by capacity and output Either alone fails. Capacity alone ("a strategist two days a week") gives the client no sense of what arrives. Deliverables alone ("eight assets a month") turns an agency into a factory and stops anybody thinking. State both: this team, this much time, producing roughly this. ## Name what is outside Production, media spend, research, third-party costs, and anything requiring a specialist not on the retained team. Agreeing this in advance converts an awkward conversation into an administrative one. ## Handle creep before it happens Small extra requests are the real problem, not large ones. Agree a mechanism: a tracked overage, a quarterly reconciliation, or an explicit allowance for ad-hoc work. Agencies that absorb everything grow resentful and quietly reduce quality; clients who are charged for every email stop asking for things they should be asking for. ## Review quarterly, properly Not a performance review. A scope review: is this still the right work, is the balance right, what should stop. Retainers decay into maintenance because nobody ever revisits the original assumption, which is the failure mode described in agency or in-house (https://colleviate.com/blog/agency-or-in-house.html). ## The exit clause Notice period, what is handed over, in what format, and who owns the accounts and source files. Write it while everybody is optimistic. It is the single most useful paragraph in the contract and the one most often left to a template. ## When not to use a retainer When the work has an end. A website, a brand refresh, a launch film. Those are projects, and billing them monthly for a year after delivery is how clients come to feel they are paying for a meeting. ## Questions **How should an agency retainer be scoped?** By capacity and output together: named people at stated seniority for a stated time, producing a described volume of work, with everything outside it listed. **How do you prevent scope creep?** Agree the mechanism before it happens: a tracked overage, a quarterly reconciliation or an explicit ad-hoc allowance. Absorbing it silently damages the work. --- # Landing pages that convert: what the evidence supports URL: https://colleviate.com/blog/landing-pages-that-convert.html Topic: Method | Published: 2026-09-08 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Landing pages convert better mainly by reducing the effort a visitor has to make. Ask only for the form fields you need for the first call, make the page fast on mobile, and remove interstitials, unprompted chat widgets and anything covering the call to action. Test with your own form on a mid-range phone over mobile data. Most conversion advice is folklore repeated with confidence. A smaller set of things reliably works, and nearly all of it is about reducing the effort a visitor has to make. A landing page has one job: let someone who arrived with a question decide, quickly, whether you are the answer. Nearly every reliable improvement comes from removing something that got in the way of that. ## The structure that holds up - Say the same thing the ad said. Message match is the most common failure. If the ad promised a price and the page opens with a brand story, the visitor has to translate, and a share of them will not. - Lead with the specific claim. What it is, who it is for, what it costs or how it is priced. Vague headlines protect nobody. - Show evidence early. Numbers, named clients, a screenshot, a real photograph of the actual thing. Stock imagery signals nothing. - One primary action, repeated down the page. Secondary actions are fine as long as they look secondary. - Answer the objection on the page. Price, contract length, timelines, who owns what. Every unanswered objection becomes an abandoned session. ## Friction worth removing Every extra form field costs completions. Ask for what you need to make the first call and nothing more; you can qualify in the conversation. Slow pages cost completions in proportion to the delay, and mobile is where the losses concentrate. Interstitials, chat widgets that open unprompted, and cookie banners covering the call to action are all self-inflicted. A mundane check that repeatedly finds money: complete your own form on a mid-range phone, on mobile data, as a new visitor. Time it. Note every moment you had to think. That list is usually the roadmap. ## What testing can and cannot tell you A/B testing needs traffic and patience. At a few hundred visits a week you cannot detect anything short of a large effect, and stopping a test the moment it looks positive manufactures results that do not repeat. Below roughly a thousand conversions a month, spend the effort on obvious fixes and qualitative evidence — session recordings, five customer interviews, the questions your sales team keeps answering — rather than on a testing programme you lack the volume to run. When you do test: one meaningful change at a time, a duration fixed in advance, whole weeks to avoid day-of-week effects, and a metric that is a business outcome rather than a click. ## The button colour question It does not matter, beyond the button being visible and obviously clickable. Button-colour case studies persist because they are cheap to run and easy to write up. The tests that produce durable gains are almost always about the offer, the proof or the effort required — the parts that need a decision from someone senior. ## Before you redesign Redesigns reset what you learned and frequently lose ground that was invisible until it was gone. Fix the specific failures first, measure, and redesign only when the page is structurally wrong rather than merely dated. If you are redesigning anyway, read how to do it without losing rankings (https://colleviate.com/blog/website-redesign-without-losing-rankings.html), and put the measurement in place before the change: see GA4 that survives scrutiny (https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html). ## Questions **What is a good landing page conversion rate?** It depends on traffic source and offer. Paid search to a considered B2B offer might convert at 2 to 5 percent, a low-commitment consumer offer far higher. Compare against your own baseline, not a published average. **How many form fields should we ask for?** The fewest that let you have a useful first conversation. Every additional field costs completions, and qualification questions are usually better asked by a person. **Do we have enough traffic to A/B test?** If you see fewer than about a thousand conversions a month, testing will rarely produce reliable answers. Use qualitative evidence and fix obvious friction instead. --- # The first ninety days with a new agency URL: https://colleviate.com/blog/first-ninety-days-with-an-agency.html Topic: Method | Published: 2026-09-08 | Updated: 2026-09-08 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Week one is access and tracking. Week two is an agreed baseline nobody disputes. Week four is a written diagnosis including what is not worth doing. First work should be live inside the first month, and the ninety-day review should be against the baseline, not against the plan. If month one is spent on ideation workshops, budget for a slow year. The shape of the first quarter predicts the relationship more reliably than the pitch did. There is a standard sequence, and departures from it are usually a warning. ## Week one: access and tracking Ad accounts, analytics, Search Console, CMS, CRM, brand assets, previous reporting. Then an immediate check of whether the measurement is sound, because every later conclusion depends on it. The checklist is in the ad account audit checklist (https://colleviate.com/blog/ad-account-audit-checklist.html). ## Week two: the baseline An agreed statement of where things stand: traffic, pipeline, cost per acquisition, conversion rates, search position, the lot. Both sides sign it. Without this, month six becomes an argument about what the numbers were at the start, and that argument has no winner. ## Week four: the diagnosis Written. What the constraint is, what the evidence says, what the agency recommends and, critically, what it recommends against. An agency that recommends everything it sells has not diagnosed anything. ## Month one: something live Not the campaign. Something: a fixed tracking issue, a rebuilt landing page, a corrected feed, a first creative test. Momentum matters, and it demonstrates the team can ship rather than only present. ## Month three: the real review Against the baseline, with the leading indicators weighted more heavily than the lagging ones, because ninety days is too short for most lagging numbers to have moved. Lead with what did not work. An agency that has learned nothing in a quarter has not tested anything. ## What should worry you No written diagnosis by week six. Reporting that only contains wins. A senior team at the pitch who have not been seen since. And a scope conversation that keeps being deferred, which usually means the scope was wrong: see scoping a retainer (https://colleviate.com/blog/scoping-a-retainer.html). ## Questions **What should happen in the first month with a new agency?** Access and a tracking check, an agreed baseline, a written diagnosis, and at least one thing shipped. Ideation workshops are not a first month. **How should the ninety-day review be judged?** Against the agreed baseline, weighting leading indicators, and opening with what did not work. Most lagging metrics will not have moved yet. --- # A GA4 setup that survives scrutiny URL: https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html Topic: Measurement | Published: 2026-09-07 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A GA4 setup survives scrutiny when it starts from a written measurement plan, not a tag. Define the handful of actions that indicate progress, with precise triggers and values, implement consent mode with denied defaults where required, and reconcile the numbers with finance once a quarter. Analytics fails quietly, so specify before you count. Most analytics implementations are installed once, never specified, and then quoted in board meetings as though they were an accounting system. Analytics does not fail loudly. It produces numbers whatever you do, and they look plausible until somebody compares them with the invoices. The fix is boring and worth it: decide what you are counting before you count it. ## Start with a measurement plan, not a tag Write down the handful of actions that indicate progress: enquiry submitted, qualified enquiry, purchase, key page reached, call started. For each, define the trigger precisely and the value if it has one. The document is a page long and it prevents the most common failure, which is eleven similar events nobody can tell apart six months later. ## Events worth defining - Lead events, separated by quality. A form submission and a qualified enquiry are different things and should not share a name. - Contact intent. Email clicks, phone clicks, WhatsApp clicks. For many Indian businesses these matter more than form fills and are routinely untracked. - Commercial milestones. Pricing page reached, calculator used, sample requested. These are the leading indicators inside the site. - Outbound clicks where a partner or marketplace completes the transaction. ## Consent, done properly If you operate in or market to places with consent requirements, implement consent mode with denied defaults and update on the visitor’s choice, rather than firing everything and hoping. Expect a share of traffic to decline analytics storage; modelling fills part of that gap and the rest is a known, disclosed limitation. Under India’s data protection regime, being able to show what you collect and on what basis is becoming an operational requirement rather than a legal nicety. ## The reports that mislead - Default channel groupings. Useful until a campaign is tagged inconsistently, after which a chunk of paid traffic appears as direct. Enforce UTM conventions in writing. - Attribution comparisons. GA4’s model and each ad platform’s model will disagree. Neither is wrong; they answer different questions, and neither is revenue truth. - Engagement rate as a quality signal. It measures time and scroll, not usefulness. - Any period spanning a tracking change without a note saying so. ## Reconcile with finance, once a quarter Take a month, pull analytics-reported conversions and revenue, and compare with the finance system. A gap is normal; an unexplained gap is a problem; an unexamined gap is how a marketing team ends up reporting numbers the CFO does not recognise. Write the reconciliation down, including the reasons for the variance, and reuse the explanation. ## Where server-side helps Server-side tagging improves data quality where browser restrictions and ad blockers bite, and gives you control over what is sent to third parties. It also adds cost and an engineering dependency. It earns its place when paid media is a large budget line or the data feeds decisions of consequence. It is rarely the right first step for a business that has not yet defined its events. Once the numbers are trustworthy, the next problem is presentation. Most reporting fails not because the data is wrong but because nobody can see the decision in it: see the dashboard that gets read (https://colleviate.com/blog/marketing-dashboard-that-gets-read.html). Related. Choosing the rest of the stack: a martech stack for small teams (https://colleviate.com/blog/martech-stack-for-small-teams.html). ## Questions **Is GA4 enough on its own?** For most businesses, yes, if it is specified properly and reconciled with the finance system. It is not an accounting record and should not be presented as one. **How much data do we lose to consent rejections?** It varies by audience and geography, commonly a noticeable double-digit share. Consent mode modelling recovers part of it; the honest approach is to disclose the limitation rather than quietly ignore it. **Should we move to server-side tagging?** Consider it once paid media is a significant budget line, or when data quality is materially affecting decisions. It is rarely the right first investment. --- # The pricing page is a sales conversation URL: https://colleviate.com/blog/pricing-page-that-sells.html Topic: Method | Published: 2026-09-07 | Updated: 2026-09-07 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Show a price or a credible range unless you have a specific reason not to. Hiding it increases enquiries and decreases qualified ones. Three tiers, an obvious recommendation, and the objections answered on the page rather than left for a sales call. It is the most visited page on most business websites and usually the least considered. People arrive at a pricing page with one question and a lot of suspicion. Everything on the page either answers the question or increases the suspicion. ## Show the price "Contact us for pricing" filters out everybody who will not tolerate a sales call to learn a number, which includes most of the people who would have bought. It also signals that the price varies by what the seller thinks you can pay, which some buyers find objectionable and all buyers notice. Where genuine custom pricing is unavoidable, publish a starting point or a range. The point is to let someone self-qualify. ## Three tiers, with a recommendation Three is enough to anchor without paralysing. Make one visibly recommended, and make the difference between tiers about a dimension the buyer understands, such as volume or users, rather than a feature matrix only you can parse. ## The enterprise tier Fine to have, provided it says who it is for and what changes, not just "contact sales". A tier with no information is a dead end that the buyer reads as "expensive and slow". ## Answer the objections on the page Contract length, what happens at renewal, what is extra, implementation time and cost, what happens to the data if you leave. Every one of these left unanswered becomes either an abandoned session or a sales call spent on administration rather than value. ## Testing with limited traffic Most pricing pages do not have the volume for reliable A/B testing. Use the questions your sales team answers most often as the roadmap instead, and treat a sharp drop between pricing page and next step as the signal. The wider method is in landing pages that convert (https://colleviate.com/blog/landing-pages-that-convert.html). ## Questions **Should we show prices on our website?** Usually yes, or a credible range. Hiding price raises enquiry volume and lowers quality, and it costs you buyers who will not call to learn a number. **How many pricing tiers should there be?** Three is usually right, with one visibly recommended and differences expressed in terms the buyer already understands. --- # Content that earns links, and content that just fills a calendar URL: https://colleviate.com/blog/content-that-earns-links.html Topic: Method | Published: 2026-09-06 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Content earns links when a competent person in your industry would send it to a colleague. Most calendar content fails because topics come from keyword tools, the research is reading what already ranks, and no real expert touches it. Pick real questions, say something a specialist could not have guessed, organise into clusters, and plan distribution. Most content programmes are a publishing habit in search of a reason. The pieces that work are usually the ones somebody senior did not want to give away. There is a straightforward test for whether a piece of content will do anything: would a competent person in your industry send it to a colleague? Almost nothing produced to fill a calendar passes that test, and almost everything that earns links does. ## Why so much of it produces nothing Three causes, usually together. The topic was chosen from a keyword tool rather than from a real question. The piece says what everyone else already said, because it was researched by reading the pages that currently rank. And nobody with actual expertise touched it, so it contains no claim a specialist could not have guessed. The result ranks nowhere, gets no links, and is presented monthly as output. It is expensive precisely because it looks like progress. ## What makes a piece worth citing - Original evidence. Your own data, a survey of your market, a teardown of twenty examples, a cost breakdown nobody publishes. This is the single most reliable way to earn links, and the most work. - A genuine position. Something arguable, held with reasons. Pieces that offend nobody get shared by nobody. - Specific usefulness. The template, the checklist, the calculation. People link to things they expect to need again. - Clarity about who it is for. A piece written for everyone is read by nobody, and ranks for nothing. ## Clusters, not calendars Search rewards depth on a subject rather than breadth across many. Pick a small number of subjects you genuinely want to be known for. For each, write the central explainer and then five to ten pieces answering the specific questions around it, all linked to each other and to the centre. Three well-built clusters beat sixty unrelated posts, and they are easier to maintain. This also matters for AI answers, which weight corroboration and structure heavily: a cluster gives an engine several consistent sources to draw on. We set that out in AEO and GEO (https://colleviate.com/blog/aeo-geo-ai-visibility.html). ## A cadence a small team can hold One substantial piece a month, properly researched, beats weekly filler. If a specialist has to spend two hours being interviewed for it, budget those two hours; the alternative is a piece with no expertise in it. Update rather than republish: a piece that ranks and is refreshed annually outperforms a new piece on the same subject, and updating is a quarter of the work. ## Distribution is not optional Publishing is roughly half the job. The other half is putting the piece in front of the people who might cite it: the journalists covering your category, the newsletter writers, the communities where your buyers argue with each other. One well-targeted email to someone who genuinely covers the subject beats a hundred templated outreach messages, which is the same lesson as the one about briefs in the brief is not the idea (https://colleviate.com/blog/the-brief-is-not-the-idea.html). ## How to tell if it is working Not by traffic in month two. Look at referring domains earned without payment, non-branded impressions on the cluster terms, the number of pieces that get cited or quoted, and whether sales can point to a piece that shortened a conversation. That last one is the reason B2B content exists at all, and it is the least measured. Related. The title decides whether the page is opened in the first place: headlines that earn the click without lying (https://colleviate.com/blog/headlines-that-earn-the-click-without-lying.html). For links inside your own site, see internal linking for topic clusters (https://colleviate.com/blog/internal-linking-for-topic-clusters.html). ## Questions **How often should we publish?** Less often than most calendars assume. One properly researched piece a month, with existing pieces updated, outperforms weekly filler on nearly every measure. **Does AI-written content rank?** It can, and it rarely earns links or citations, because it contains nothing a model could not already produce. Use it for drafting and structure, not for the claims. **How long should a post be?** As long as the question needs. Length correlates with ranking mostly because thorough answers tend to be longer, not because word count is a ranking factor. --- # Customer research when you cannot afford research URL: https://colleviate.com/blog/customer-research-on-a-small-budget.html Topic: Method | Published: 2026-09-06 | Updated: 2026-09-06 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Eight to twelve conversations with recent buyers and recent losses will surface almost every pattern that matters. Ask what they did and what happened, never what they would do, and mine your own reviews and sales call recordings before paying anyone. The expensive version is a survey of a thousand people answering questions they have not thought about. The cheap version is better. Most marketing decisions are made on assumptions about buyers that nobody has checked in years. Checking them is far cheaper than people assume and the constraint is almost always willingness rather than budget. ## Who to talk to Recent buyers, because the decision is fresh. Recent losses, because they will tell you what the winner had. And people who nearly bought and then did nothing, who are the most informative group and the one nobody calls. ## How many Eight to twelve per segment. Patterns repeat faster than people expect, and by the ninth conversation you will usually be hearing things you have already heard. That is the signal to stop, not a reason to doubt the sample. ## Ask about behaviour, not opinion "What would make you switch" produces fiction. "Tell me what happened the day you decided to look" produces evidence. Walk them through the actual sequence: what triggered it, who else was involved, what they searched, who they shortlisted, what nearly stopped them. ## Free sources you already have - Your own reviews and your competitors', read properly rather than counted. - Sales call recordings, where the objection you keep losing to is stated out loud weekly. - Support tickets, which contain the gap between what you promised and what happened. - Search queries from Search Console, which show the language people actually use. ## What to do with it Write the five sentences you heard repeatedly, in the customers' own words, and put them where the work gets briefed. That is usually enough to change a campaign, and it is the evidence layer underneath the outside-in brand audit (https://colleviate.com/blog/outside-in-brand-audit.html). ## Questions **How many customer interviews are enough?** Eight to twelve per segment. Patterns repeat quickly, and hearing the same thing again is the signal to stop rather than a reason to doubt the sample. **What is the biggest mistake in customer research?** Asking what people would do. Ask what they actually did and what happened; stated intention is close to worthless. --- # How to redesign a website without losing your rankings URL: https://colleviate.com/blog/website-redesign-without-losing-rankings.html Topic: Method | Published: 2026-09-05 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** To redesign a website without losing rankings, map every old URL one to one to its nearest equivalent with permanent redirects, never mass-redirect to the homepage, keep the words that ranked, carry across structured data, canonical tags, alt text and internal links, keep staging out of the index, and watch the first fortnight closely. Traffic collapses after a relaunch for boring, preventable reasons. Every one of them is on a checklist somebody did not run. A redesign changes three things at once: the addresses of your pages, the words on them, and the speed at which they load. Search engines are sensitive to all three. Handle them deliberately and a migration costs you nothing. Handle them casually and you can lose half your organic traffic in a fortnight, then spend two quarters earning it back. ## Before anything is designed - Export a full crawl of the current site. Every URL, with its traffic, rankings, backlinks and conversions. This is the inventory you will protect. - Identify the pages that matter. Usually a minority of URLs carry most of the value. Those pages get individual attention; the rest can be handled in bulk. - Record your baselines. Organic sessions, impressions, rankings by cluster, and conversion rate, for at least the previous twelve months. Without this you will not be able to prove what happened. ## The mapping Every old URL needs a destination: the same page, its nearest equivalent, or a deliberate decision to remove it. Map one to one wherever a real equivalent exists. Redirecting everything to the homepage is the classic error; search engines treat a mass of homepage redirects as soft errors and the link equity evaporates. Use permanent redirects for anything staying. Keep the map as a spreadsheet you can test against, not as a set of rules somebody remembers. ## Do not throw away the words Redesigns usually cut copy, because less text looks cleaner. Pages that ranked did so partly because of what they said. If a page must get shorter, make sure the substance survives somewhere and that the headings still describe what the page is about. Keep the page titles and descriptions unless you have a reason to change them. ## Carry the technical work across Structured data, canonical tags, hreflang if you have it, image alt text, and the internal linking (https://colleviate.com/blog/internal-linking-for-topic-clusters.html) that connected your clusters together. These are invisible in a design review and are frequently rebuilt from nothing. Performance matters too: a beautiful site that loads slowly on a mid-range phone will lose both rankings and conversions. ## Staging hygiene Block the staging site from indexing, then remember to unblock the live one. A live site launched with a blanket disallow, or with a stray noindex tag left in the template, is among the most common and most expensive mistakes in this whole list. Check it on day one, not on the day traffic drops. ## The first fortnight - Submit the new sitemap and watch index coverage daily for two weeks. - Crawl the live site and fix every redirect chain and broken internal link. - Compare impressions by page against the pre-launch baseline, not against last week. - Confirm analytics and conversion tracking fire on the new templates. This is nearly always broken on launch day: see GA4 that survives scrutiny (https://colleviate.com/blog/ga4-setup-that-survives-scrutiny.html). - Expect a dip of a few weeks as pages are recrawled and re-evaluated. A dip is normal; a cliff is a bug. If the redesign is part of a wider brand change, the stakes are higher still, because the domain and the name may move together. That is a different decision with its own economics: see rebranding, when it is worth it (https://colleviate.com/blog/rebranding-when-its-worth-it.html). ## Questions **How much traffic will we lose after a redesign?** With a proper migration, a dip of two to four weeks and then recovery. Losses that persist beyond a couple of months usually trace to redirect mapping, removed content, or an indexing block left in place. **Should we change URLs during a redesign?** Only where there is a real reason. Changing URLs for tidiness costs equity and buys nothing your visitors will notice. **Do we need to keep old blog posts?** Keep anything with links, rankings or conversions. Pages with none of those can be removed, provided the removal is deliberate and the URLs are handled. --- # What to review weekly, monthly and quarterly URL: https://colleviate.com/blog/marketing-reporting-cadence.html Topic: Measurement | Published: 2026-09-05 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Weekly for operational metrics a team can act on within the week. Monthly for channel performance and spend decisions. Quarterly for strategy, brand indicators and anything with a long lag. Nothing that moves slowly should be looked at daily, because reacting to noise is worse than not looking. Half of all bad marketing decisions come from reading a number more often than it can meaningfully change. Every metric has a natural frequency: the rate at which it can move for a real reason. Reading it faster than that means reading noise, and reading noise produces action, which is how accounts get destabilised by people trying to help. ## Weekly: operational Spend pacing, lead volume, cost per lead, response time on enquiries, anything broken. The test for inclusion is whether the team can do something about it this week. Keep it to a short list and a fifteen-minute meeting. ## Monthly: performance Channel performance, blended efficiency, cost per acquisition by cohort, content and search progress, creative test results. This is where budget moves between channels and where the commentary matters more than the charts, as argued in the dashboard that gets read (https://colleviate.com/blog/marketing-dashboard-that-gets-read.html). ## Quarterly: strategy Share of search and brand indicators, pipeline quality, payback period, whether the diagnosis in the annual plan still holds, and what to stop. The quarterly meeting should be able to change the plan; the monthly one should not. ## Never daily Rankings. Brand search. Conversion rate on low-volume pages. Anything where a single day's variance exceeds a month's trend. If somebody insists on a daily dashboard, give them spend and leads only, because those are the two that can genuinely need same-day attention. ## The discipline that makes it work Write the commentary before the meeting and circulate it. Meetings spent reading numbers aloud produce no decisions. Meetings spent arguing about an interpretation that everybody has already read produce several. Related. What the board reads: data-driven marketing, defined (https://colleviate.com/blog/data-driven-marketing-definition.html). ## Questions **How often should marketing performance be reviewed?** Weekly for operational metrics, monthly for channel performance and budget, quarterly for strategy and brand indicators. **What should not be tracked daily?** Anything where one day of variance exceeds a month of trend: rankings, brand search, and conversion rates on low-volume pages. --- # The marketing dashboard that actually gets read URL: https://colleviate.com/blog/marketing-dashboard-that-gets-read.html Topic: Measurement | Published: 2026-09-04 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A marketing dashboard gets read when it is built around the decisions its reader can make: spend more, spend less, move budget, change the offer, stop something. Every element should inform one of those; anything else is decoration. Keep the board view to one page with honest comparisons, and automate the collection, not the interpretation. A dashboard nobody opens is not a reporting problem. It is a sign that the report was built to describe activity rather than to support a decision. Most marketing reporting answers a question nobody asked: what did we do last month. The people receiving it want to know something else: is this working, what changed, and what are you doing differently as a result. ## Build it around decisions Start from the decisions your reader can make: spend more, spend less, move budget between channels, change the offer, stop something. Every element on the dashboard should inform one of those. Anything that informs none of them is decoration, however interesting it is. ## What a board actually needs - Pipeline or revenue from marketing, against plan, with the trend. - Blended efficiency: total revenue against total marketing spend, which cannot be gamed by attribution. - Cost per acquisition by cohort and payback period, so the quality of growth is visible. - One leading indicator, because everything else lags. Share of search (https://colleviate.com/blog/share-of-search-leading-indicator.html) is the cheapest credible one. - What changed and what we did about it. Three sentences of prose, which is the part people actually read. That is one page. Everything else belongs in the channel review, where the people who can act on it are in the room. ## What to leave out on purpose Impressions and reach, unless you are running an awareness campaign and have agreed what they are for. Social followers. Email list size. Platform-reported returns that contradict the finance system. And any metric that only ever goes up, because a number that cannot fall carries no information. ## Make the comparison honest Month against the same month last year, not against last month, for anything seasonal. Show the plan alongside the actual. Note the tracking changes on the chart itself; a step change caused by a new tag is not a result, and somebody will quote it as one within a quarter. ## The monthly review Lead with what did not work. It sets the tone that the meeting is for learning rather than defending, and it front-loads the only part with decisions in it. Then the numbers, briefly, because they were circulated beforehand and should not be read aloud. Then next month, stated as choices with a rationale. The agency or team that opens with its wins every month is training the room to discount the report. The one that opens with the failed test and what it changed gets believed when it does claim a win. ## Automate the collection, not the interpretation Pull the numbers automatically, by all means. The commentary has to be written by someone who understands the business, because the value is in the sentence explaining why the line moved, not in the line. That is also why fully automated reporting tends to get ignored: it produces charts without an argument, and nobody makes a decision from a chart alone. Related. A flattering average can hide a weakening base: how blended retention rises while the business gets worse (https://colleviate.com/blog/how-blended-retention-rises-while-the-business-gets-worse.html). ## Questions **How many metrics should a marketing dashboard have?** For a board, five or six, on one page, with a short written commentary. Channel teams need more detail, in a different document. **How often should we report?** Monthly to the board, weekly within the team for operational metrics. Quarterly is too slow to change anything and daily reporting encourages reacting to noise. **Who should write the commentary?** Someone who understands the commercial context, not the tool. The interpretation is the deliverable; the charts are the evidence for it. --- # When to fire your agency, and when the problem is you URL: https://colleviate.com/blog/when-to-fire-your-agency.html Topic: Method | Published: 2026-09-04 | Updated: 2026-09-04 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Switch when the agency stops arguing with you, when the senior people who pitched have vanished, or when reporting has become advocacy. Do not switch when the brief keeps changing, approvals are impossible or the budget cannot support the ask, because those problems will follow you. Agency changes are expensive, disruptive and frequently solve nothing, because half the time the dysfunction was not on the agency side. Before changing, work out which kind of failure you have. The symptoms look identical from the inside and the remedies are opposites. ## Signals that it is the agency - They have stopped arguing with you. The most reliable indicator. An agency that agrees with everything has decided the relationship is about retention rather than results. - The senior people have disappeared. Pitched by directors, serviced by juniors, which is normal in moderation and fatal past a point. - Reporting is advocacy. Every month a win, no month a failure, no test that did not work. - Repeated avoidable errors, particularly in tracking, billing or compliance. - No point of view. They execute what you ask and nothing more, which means you are paying agency rates for a production service. ## Signals that it is you - The brief has changed three times this quarter. - Work dies in approvals, or a senior objection arrives at the final stage every time. - The budget cannot support the outcome being demanded, and everyone is pretending otherwise. - Nobody internally owns the relationship, so the agency is managed by whoever is least busy. - The agency has told you the constraint is elsewhere and nobody acted on it. Changing agencies will not fix any of those. It will reset six months of context and reproduce the same outcome with new names. ## The conversation to have first A direct one, with the senior person, about what is not working, with specifics and a date to review it. Agencies can almost always fix a stated problem. They cannot fix dissatisfaction nobody has named. ## Running the transition If you do switch: get the handover in writing, keep account ownership and historical data, overlap by a fortnight where possible, and give the incoming team a baseline rather than a grievance. Then run the first ninety days properly: see the first ninety days with a new agency (https://colleviate.com/blog/first-ninety-days-with-an-agency.html). ## Questions **What is the clearest sign to change agency?** They have stopped disagreeing with you. An agency protecting the relationship rather than the results has stopped being useful. **What should we check before switching agencies?** Whether the brief keeps changing, whether approvals kill the work, and whether the budget supports the ask. Those problems move with you. --- # Agency or in-house: the honest comparison URL: https://colleviate.com/blog/agency-or-in-house.html Topic: Method | Published: 2026-09-03 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Own what needs continuity and context: the strategy, customer understanding, brand decisions, the relationship with sales and the data. Buy what needs range or capacity: creative, production, specialist channel work and an occasional outside read. The common failure is the inverse, hiring in-house to save on execution and outsourcing strategy, so nobody owns the direction. The comparison is usually framed as cost. It is really about which capabilities you need continuously and which you need occasionally. Marketing needs perhaps a dozen distinct capabilities: positioning, planning, creative direction, copy, design, film, media buying, SEO, lifecycle, analytics, production management, and someone to hold it together. Almost no company needs all of them full time, and almost no individual has more than three of them at a serious level. ## The real cost of in-house A competent mid-level marketing manager in a metro costs more than the salary: add recruitment, equipment, tools, management time, and the months before they are productive. Then add the specialists they will need. A single hire who is genuinely strong at paid media, SEO, design, film and analytics does not exist; hiring as though they do produces a generalist doing five jobs at an average standard and burning out doing it. What in-house buys that an agency cannot: context, availability, institutional memory, and someone whose only client is you. ## The real cost of an agency A retainer is visible, which makes it easy to scrutinise and easy to cut. Add the time your team spends briefing and reviewing, which is genuine cost and routinely omitted. Add the risk that the agency’s attention follows its largest account. What an agency buys that a hire cannot: a range of specialisms on demand, pattern recognition from other categories, production capacity that scales, and the ability to stop without a redundancy process. ## The split that usually works Own the things that need continuity and context: the strategy, the customer understanding, the brand decisions, the relationship with sales, and the data. Buy the things that need range or capacity: creative, production, specialist channel work, and the occasional outside read on whether you are fooling yourself. The common failure is the inverse. Companies hire in-house to save money on execution, then outsource strategy to the agency, and end up with an operating model where nobody owns the direction and everybody owns a deliverable. ## Thresholds worth knowing - Below roughly Rs 3 lakh a month of total marketing spend, a single senior generalist plus freelancers usually beats both a full agency retainer and a team. - Between that and about Rs 15 lakh, a hybrid tends to win: one or two in-house owners, an agency for range. - Above that, bringing high-frequency work in-house starts to pay, while production and specialist campaigns usually stay outside. ## The failure modes In-house: the team becomes a service desk for other departments, strategy disappears under requests, and the work drifts towards whatever is easiest to approve. Agency: the relationship settles into a monthly rhythm nobody questions, the work becomes maintenance, and the agency stops arguing with you, which is exactly when it stops being useful. Both failure modes are prevented the same way: a written definition of what is being bought and a quarterly review that asks whether it is still the right thing. On the agency side, that review is far easier when you agreed the scope properly at the start: see what a marketing agency should cost (https://colleviate.com/blog/what-a-marketing-agency-should-cost.html). Related. Who to hire first: marketing team structure (https://colleviate.com/blog/marketing-team-structure.html); if you go external, how to choose a digital marketing agency (https://colleviate.com/blog/how-to-choose-a-digital-marketing-agency.html). ## Questions **Is an in-house team cheaper than an agency?** Rarely, once you count recruitment, tools, management time and the specialists a single hire cannot replace. It buys context and availability, which can be worth more than the saving. **What is the first marketing hire?** Someone senior enough to own direction and commercially literate enough to argue with sales and finance. Hiring an executor first leaves nobody deciding what to execute. **Can we run both?** That is the most common working model above modest spend: own strategy, customer understanding and data, and buy creative, production and specialist channel work. --- # What video production actually costs in India, and what drives the number URL: https://colleviate.com/blog/video-production-cost-india.html Topic: Creative | Published: 2026-09-02 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Video production quotes in India for the same film can differ fourfold, and both can be honest, because the variance comes from decisions the brief made or left open: the number of shoot days, which is the biggest lever, locations, talent and usage rights, post-production complexity and versioning. Settle them early and treat the agreed treatment as the contract. Two quotes for the same film can differ fourfold, and both can be honest. The variance is almost entirely in decisions the brief either made or left open. A production quote is a stack of day rates, kit, people, locations and post. Almost everything that moves the total is decided before anyone picks up a camera, which is why the brief is the real budget document. ## What is in a quote - Pre-production. Treatment, storyboard, casting, location recce, scheduling, permissions. Underfund this and you pay for it on shoot day at a much higher rate. - Crew. Director, DOP, gaffer, sound, assistants, stylist, and a producer who keeps the day on schedule. - Kit and locations. Camera package, lighting, grip, studio or location fees, permissions, transport. - Talent. Fees plus usage rights, which are a separate negotiation and frequently the line that surprises people later. - Post. Edit, colour grade, sound design, music licensing, motion graphics, and versioning for each placement. ## The decisions that move the number most Number of shoot days. The single biggest lever. Every day carries the whole crew and kit, so a script that can be shot in one day rather than two is worth more than any negotiation. Locations. A controlled studio is usually cheaper than three real locations with travel, permissions and weather risk. Talent and usage. A known face is not just a fee; it is a fee multiplied by territory and duration. Two years of all-media usage costs a multiple of six months of digital. Post complexity. Motion graphics, visual effects and elaborate sound design are open-ended in a way that shooting is not. Versioning. Ten cut-downs in five aspect ratios is a real edit workload, and it should be in the brief rather than requested after delivery. ## Where you can save safely Shoot several things at once. If you need a brand film, product stills and social cut-downs, plan them into the same days with the same crew. This is by a distance the most effective saving available, and it requires only that the brief exists early enough. Cast real people where the story allows it. Use fewer locations and light them better. Shoot for the edit you have actually planned rather than covering every possibility. And agree the number of revision rounds in writing, because unlimited revisions are paid for somewhere, usually in the next quote. ## Where saving will show Sound. Bad audio reads as amateurish faster than any other flaw, and fixing it in post is expensive when it is possible at all. Grading, which is what makes footage look like a film rather than a recording. Enough pre-production time to have solved the problems before the day. And the producer, who is the person preventing the expensive chaos you will otherwise pay for in overtime. ## Treat the treatment as the contract Everything above argues for the same discipline: decide the look, the shot list, the locations, the talent, the deliverables and the usage before the shoot is booked. A quote against a vague brief is a guess, and the gap between the guess and reality is always settled in the client’s favour on paper and against them on screen. Our view of what a brief owes the work is in the brief is not the idea (https://colleviate.com/blog/the-brief-is-not-the-idea.html). Related. What to brief before the shoot: a photography direction brief (https://colleviate.com/blog/photography-direction-brief.html) and short-form video that works (https://colleviate.com/blog/short-form-video-that-works.html). ## Questions **How much does a corporate film cost in India?** A simple single-day corporate film with a small crew commonly runs a few lakh rupees. Multi-day campaign films with talent, locations and full post run to several times that. The shoot-day count and usage rights drive most of the variance. **Why are two quotes for the same film so different?** Usually because they assume different shoot days, crew sizes, talent usage and post scope. Ask each supplier to state those assumptions and the comparison becomes possible. **Can we reuse footage across campaigns?** Only within the usage rights you bought, for talent, music and stock. Agree the territory and duration up front; extending later is always more expensive than buying it at the start. --- # Social media marketing that sells, not just posts URL: https://colleviate.com/blog/social-media-that-sells.html Topic: Creative | Published: 2026-09-01 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Social media sells when it starts from what you want to be known for, and to whom, rather than a posting schedule. Organic reach is rationed, so choose fewer platforms based on where your buyers argue, post three genuinely good pieces rather than daily filler, and run organic and paid together. Report saves, shares, profile visits, qualified enquiries and branded search, not followers. Most social strategies are a posting schedule with a mission statement attached. The ones that work start from a much narrower question: what do we want to be known for, and to whom. Organic social is not a distribution channel any more. Reach is rationed, feeds are ranked by engagement rather than by who follows you, and posting into that system without something worth stopping for is a form of expensive silence. ## Choose fewer platforms Being present on five platforms with one person is a guarantee of mediocrity on all five. Choose based on where your buyers argue with each other, not where the audience is largest. For most B2B businesses that means LinkedIn and, increasingly, YouTube. For consumer brands it usually means Instagram and short video. For everyone, it means fewer than the plan currently says. ## What earns reach - A specific point of view, repeated. Accounts that stand for something get recognised in the feed; accounts that post whatever is topical do not. - Native format. Content made for the platform outperforms repurposed assets, consistently and by a wide margin. - Faces and voices. People engage with people. Not necessarily the leadership team, but not a logo either. - Usefulness or entertainment. The two reasons anyone shares anything. Announcements are neither. ## Frequency, honestly Three good posts a week beats daily filler, provided the three are genuinely good. Consistency matters, but consistency at low quality teaches the algorithm and the audience the same lesson. Decide a cadence your team can hold in a busy month, not the one that looks ambitious in a plan. ## Organic and paid are not separate Treat organic as the testing ground and paid as the amplifier. Post ten ideas, watch which three earn attention without support, then put budget behind those. This is cheaper and more reliable than writing ads in a vacuum, and it makes the creative argument evidential rather than aesthetic. The channel logic behind the spend is in Google Ads or Meta Ads (https://colleviate.com/blog/google-ads-vs-meta-ads.html). ## The metrics worth reporting Not followers. Follower count is the vanity metric that refuses to die, and it has almost no relationship to revenue. Report saves and shares, which indicate value; profile visits and click-throughs, which indicate intent; qualified enquiries attributed to social, however imperfectly; and branded search volume, which often moves before anything else does. ## Community is slower and worth more Replying properly, showing up in other people’s comments, and answering the awkward questions in public builds something a campaign cannot. It is unglamorous, it does not fit a quarterly plan, and in categories with long consideration cycles it is frequently the highest-return activity available. It also compounds, which is the argument we make more generally in audacity and tenacity (https://colleviate.com/blog/audacity-and-tenacity.html). ## What to stop doing Posting on every platform because it exists. Chasing trends that have nothing to do with your business. Announcing your own news as though it were content. Engagement-bait questions. And rebuilding the whole approach every time a new format appears, which is how accounts end up with five years of posts and no recognisable identity. Related. On format and community, see short-form video that works (https://colleviate.com/blog/short-form-video-that-works.html), community management in India (https://colleviate.com/blog/community-management-india.html) and influencer marketing that works (https://colleviate.com/blog/influencer-marketing-that-works.html). ## Questions **How often should we post on social media?** Three strong posts a week beats daily filler on nearly every measure. Choose a cadence your team can sustain in a busy month. **Should we be on every platform?** No. Two done properly outperform five done thinly. Choose by where your buyers already are, not by audience size. **Do follower counts matter?** Very little. Saves, shares, profile visits and enquiries carry information about intent; follower count mostly reflects history. --- # Influencer marketing that works, and how to spot the money pit URL: https://colleviate.com/blog/influencer-marketing-that-works.html Topic: Culture | Published: 2026-08-31 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Influencer marketing works when creators are chosen by who is actually watching, not by follower count. Check audience data and recent post performance, read the comments for signs of a real audience, agree pricing and usage rights, and brief the outcome rather than a script. Measure with unique codes per creator, watch time and saves, and a holdout market for anything substantial. Creator marketing fails in predictable ways, and nearly all of them trace to choosing the creator by follower count rather than by who is actually watching. The pitch for influencer marketing is trust at scale: an audience that already believes someone, borrowed briefly. The failure mode is the same sentence with the trust removed, which is what happens when the match is wrong or the brief is a script. ## Choose by audience, not reach A creator with 30,000 followers who are precisely your buyer is worth more than one with a million who are not. Ask for audience demographics, geography and the last five posts’ performance, not just the follower number. Then check the comments: real audiences argue, ask questions and reference earlier posts. Generic praise in volume is a warning. The same discipline applies as in casting for a film, which we set out in casting for evidence (https://colleviate.com/blog/casting-for-evidence.html): fit with the audience beats fame, almost every time. ## Spotting inflated numbers - Engagement rate far above category norms, especially with low saves and shares. - Follower spikes that do not correspond to a post or a moment. - Audience geography that does not match the language or the market. - Comment patterns that repeat across unrelated posts. ## Pricing and rights Rates vary enormously and are negotiable. What is not negotiable after the fact is usage: if you want to run the content as a paid ad, or use it on your own channels, or keep it for a year, agree that in the contract at the start. Buying usage retrospectively costs a multiple of buying it up front, and this is the most common budget surprise in creator work. Whitelisting — running ads from the creator’s own handle — usually outperforms the same creative on the brand account, because the source is the thing the audience trusts. It also needs to be agreed in writing. ## Brief the outcome, not the script Creators know their audience better than your brand team does. Give them the proposition, the claims they must not make, and the things that are legally required. Then let them write it. Campaigns that read like a press release delivered by a stranger perform badly and damage the creator, which is why the good ones stop working with brands that insist. ## Disclosure is not optional In India, paid partnerships must be disclosed clearly and up front under the ASCI guidelines and consumer protection rules, and the liability is not only the creator’s. Insist on it. A campaign that hides a commercial relationship is a reputational risk to the brand, not just to the person paid to post. ## Measuring it honestly Unique codes and links per creator, so you can separate them. Watch time and saves rather than likes. And, for anything substantial, run a holdout: two comparable markets, creator activity in one, nothing in the other. Reported results from creator campaigns suffer from the same attribution problems as everything else, for the reasons in what ROAS actually tells you (https://colleviate.com/blog/performance-marketing-roas-reality.html). ## When it is the wrong tool For a considered, high-price, long-cycle purchase, a creator post rarely closes anything; it may still earn awareness among the right people. For a category nobody understands yet, education is usually a better use of the money. And for a brand with an unresolved product problem, the campaign simply makes the problem better known. ## Questions **Do micro creators outperform large ones?** Often, on engagement and cost per outcome, because their audiences are narrower and more trusting. Large creators buy reach, which is a different job. **What should an influencer campaign cost?** Rates range from a few thousand rupees for a nano creator to several lakhs for a well-known one. Usage rights and exclusivity change the number more than follower count does. **Do we need to disclose paid partnerships in India?** Yes. Disclosure must be clear and prominent under ASCI guidelines and consumer protection rules, and the brand carries responsibility alongside the creator. --- # D2C growth after the first crore: where the model usually breaks URL: https://colleviate.com/blog/d2c-growth-after-the-first-crore.html Topic: Growth audits | Published: 2026-08-30 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** D2C brands often stall after the first crore because early growth came from a narrow, cheap pocket of demand that does not survive scaling. Work out true contribution per order and break-even ROAS first, then build repeat purchase, which is the whole business. When performance efficiency stops improving, the constraint is that nobody has heard of you, and brand has to enter. The first crore of revenue proves people will buy. It proves almost nothing about whether the business works at ten times the size, and the gap between those two is where most brands stall. Early D2C growth is usually powered by a narrow, cheap pocket of demand: a founder’s network, one product that photographs well, a period of favourable ad costs. Scaling means leaving that pocket, and everything that was comfortably profitable inside it becomes marginal outside it. ## The four things that break, in order - Acquisition cost rises. The cheapest audience is exhausted first. Costs climb as you reach further, and the campaigns that looked efficient at Rs 3 lakh a month rarely hold at Rs 15 lakh. - Contribution margin turns out to be thinner than assumed. Shipping, returns, packaging, payment fees, discounts and COD reconciliation. Many brands discover their real per-order contribution only when volume makes the leak visible. - Repeat rate does not arrive. The model assumed a second purchase that the category may not support, or that nobody has done the work to earn. - Marketplace dependence. Volume moves to a platform that owns the customer relationship, and the brand becomes a supplier with a logo. ## Fix the unit economics before the funnel Work out the true contribution per order, after every cost that varies with the order. Then compute break-even ROAS from it. Plenty of brands scale for a year against a target that was structurally loss-making, then blame the platform. Once the true number is known, some decisions become obvious: raise the average order value, cut the discount, change the shipping threshold, or drop the product that never contributed. ## Repeat is the whole business In most D2C categories, the first order is bought at roughly break-even and the profit lives in the second and third. That makes retention a growth lever, not a loyalty programme. The practical work is unglamorous: replenishment timing, post-purchase communication that is useful rather than promotional, and a product good enough to be worth reordering. We cover the mechanics in email and lifecycle marketing (https://colleviate.com/blog/email-and-lifecycle-marketing.html). If the category genuinely does not repeat — a mattress, a wedding purchase — then the model has to work on the first order alone, which means either a much higher margin or a much lower acquisition cost. Pretending otherwise is the most expensive form of optimism in the sector. ## Where brand enters At about this point, performance efficiency stops improving no matter who runs the account, because the constraint is no longer targeting. It is that nobody has heard of you. Brands that invested earlier in being known buy the same click more cheaply and convert it better. That is the argument in brand versus performance (https://colleviate.com/blog/brand-versus-performance-split.html), and this is the stage where ignoring it becomes visible in the numbers. ## The order we would work in - True contribution per order and per product. - Retention and repeat mechanics. - Conversion rate on existing traffic. - Channel mix and creative volume. - Brand investment, held for at least four quarters. Most brands do this list backwards, starting with more ad spend, because it is the only step that can be taken on a Monday morning without anybody else’s agreement. ## Questions **What is a healthy CAC to LTV ratio for D2C?** A frequently cited benchmark is three to one, but it is meaningless without a payback period. A three-to-one ratio realised over four years is a cash flow problem, not a healthy business. **Should a D2C brand sell on marketplaces?** Usually yes, for volume and discovery, with a clear view that the customer relationship belongs to the platform. Build owned channels in parallel rather than after dependence sets in. **When should a D2C brand start brand marketing?** Once acquisition costs are rising on flat volume and conversion improvements have been exhausted. That is the signal that the constraint is awareness rather than targeting. --- # B2B lead generation that survives a sales review URL: https://colleviate.com/blog/b2b-lead-generation-india.html Topic: Method | Published: 2026-08-29 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** B2B lead generation survives a sales review when marketing and sales first agree, in writing, what counts as a qualified lead: company profile, role, trigger and action taken, with sales feedback on every rejected lead. Then report pipeline created and influenced by source, conversion from qualified lead to opportunity, sales cycle length, and what sales rejected and why. Most B2B lead generation arguments are really a disagreement about the definition of a lead, conducted through numbers neither side believes. Marketing reports a rising lead count. Sales reports that the leads are unusable. Both are telling the truth, because they are counting different things, and no dashboard will resolve a definitional disagreement. ## Start with the definition Write down, jointly, what counts as a qualified lead: company profile, role, trigger, and the action taken. Then agree what marketing owes sales, what sales owes marketing in return — specifically, feedback on every rejected lead within a fixed window — and review it monthly. This single document resolves more B2B friction than any tool purchase. ## The channels that actually work for long cycles - Search for the problem, not the product. Buyers search symptoms long before they search categories. That is where the winnable, high-intent queries are. - Content with genuine expertise, because B2B buying is a committee decision and content is what circulates inside the committee when you are not in the room. - LinkedIn, for reach among named accounts rather than for volume. It is expensive per click and often cheap per qualified conversation. - Events and communities, which convert poorly on a spreadsheet and disproportionately well in reality. - Referral and partner routes, the highest-converting channel in most B2B businesses and the least systematically worked. ## The gated content question Gating everything maximises form fills and minimises reach. Ungating everything maximises reach and gives you fewer names. The workable compromise: ungate the thinking, gate the tools. Publish the argument freely so it circulates and gets cited, and ask for details only where the visitor gets something operationally useful in return — a template, a calculator, a benchmark set. Volume of leads falls, quality rises, and sales stops complaining. ## LinkedIn without the spam Automated connection sequences work until everybody uses them, which was several years ago. What still works: a company page that publishes something worth reading, a small number of subject-matter people posting in their own voice, and sales people who comment usefully on other people’s posts before they ever send a message. It is slower and it does not fit a weekly activity target, which is precisely why it remains effective. ## Content that shortens deals The highest-return B2B content is rarely top-of-funnel. It is the material that removes friction late: the implementation detail, the security and compliance answer, the honest comparison, the pricing logic. Ask your sales team which three questions they answer in every deal, then publish the answers properly. This is the fastest available win in most B2B content programmes: see content that earns links (https://colleviate.com/blog/content-that-earns-links.html). ## What to report Pipeline created and pipeline influenced, by source. Conversion rate from qualified lead to opportunity. Sales cycle length by source, which is where good marketing shows up before revenue does. And a note on what sales rejected and why, because a lead count with no rejection data is a number nobody in the room believes. Related. When the sale is made by someone else, see measuring a business that sells through dealers, distributors and referrers (https://colleviate.com/blog/measuring-a-business-that-sells-through-others.html). ## Questions **How long does B2B lead generation take to work?** Expect two to three quarters before pipeline data is readable, longer in categories with annual budget cycles. Leading indicators move earlier: impressions on problem-led queries, content circulating inside accounts, inbound from named targets. **Should B2B content be gated?** Ungate the thinking so it circulates and gets cited; gate the tools where the visitor gets something operationally useful. Gating everything trades reach for form fills. **Is LinkedIn advertising worth it for B2B?** It is expensive per click and often reasonable per qualified conversation, particularly against a named account list. Judge it on pipeline, not on cost per lead. --- # Email and lifecycle marketing: the channel everyone underrates URL: https://colleviate.com/blog/email-and-lifecycle-marketing.html Topic: Measurement | Published: 2026-08-28 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Email and lifecycle marketing is underrated because it is unglamorous, yet in most businesses it returns more per rupee than anything else: you own the audience and sends cost almost nothing. Build behaviour-triggered flows before campaigns, keep the list clean rather than large, and measure clicks, conversion per send, revenue per recipient and engaged subscribers, not open rates. It is unfashionable, it has no agency glamour attached, and in most businesses it quietly returns more per rupee than anything else in the plan. Owned channels have one structural advantage nothing else can match: you are not renting the audience. No auction sets the price, no algorithm change removes your reach, and the cost per message is close to nothing. The reason it stays underused is that it is unglamorous and it requires someone to maintain it. ## Build the flows before the campaigns Automated flows triggered by behaviour outperform broadcast sends by a wide margin, because they arrive when the person is already thinking about you. Build them in this order: - Welcome. What you do, what to expect, and the single most useful thing you have. This is when attention is highest and most brands send a discount code instead. - Abandoned cart or abandoned enquiry. The highest-return flow in nearly every business, and frequently the only one anybody builds. - Post-purchase. Useful rather than promotional: how to get the most from it, what to expect, when to reorder. - Replenishment or renewal, timed to the actual consumption cycle rather than to a monthly calendar. - Win-back, with a genuine reason to return rather than an escalating series of discounts that trains people to wait. ## List health beats list size A large list full of people who never open is worse than a small engaged one: it damages deliverability, so your messages stop reaching the people who do want them. Clean it. Remove persistent non-openers, authenticate your sending domain properly, and never buy a list, which is both ineffective and, for personal data in India and most markets, legally hazardous. ## What to measure now Open rates became unreliable once mail clients started pre-fetching images, and using them as a headline metric now measures the mail client. Use click-through rate, conversion rate per send, revenue per recipient, unsubscribe rate as a quality signal, and the trend in engaged subscribers over ninety days. The last one is the health of the asset. ## Frequency Send as often as you have something worth sending, which is a real constraint rather than a platitude. Weekly works if the content earns it. Monthly is fine. The failure is a fixed cadence filled with whatever is available, which trains subscribers to ignore you and then to leave. ## WhatsApp, carefully In India, WhatsApp frequently outperforms email on engagement and is far less forgiving of misuse, because it feels personal. Use it for transactional and genuinely timely messages, honour opt-outs immediately, and keep promotional volume low. A channel people cannot escape is a channel they will block. ## Where it connects to everything else Lifecycle marketing is what makes paid acquisition affordable: if the second purchase is reliable, you can pay more for the first. That relationship is the whole economic argument in D2C growth after the first crore (https://colleviate.com/blog/d2c-growth-after-the-first-crore.html), and it is the reason retention work usually belongs ahead of another increase in ad spend. ## Questions **How often should we send marketing emails?** As often as you have something worth sending. A fixed cadence filled with filler trains people to ignore you, and then to unsubscribe. **Are open rates still useful?** Not as a headline metric, since mail clients pre-fetch images. Use clicks, conversions, revenue per recipient and the trend in engaged subscribers. **Which flow should we build first?** Abandoned cart or abandoned enquiry, then welcome. Those two carry most of the available return in nearly every business. --- # Rebranding: when it is worth it, and when it is a very expensive distraction URL: https://colleviate.com/blog/rebranding-when-its-worth-it.html Topic: Creative | Published: 2026-08-27 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** A rebrand is worth it when a business fact has already changed, so the recognition you paid for describes something you no longer are. Boredom, a competitor's rebrand, a dated logo or a new CMO are not reasons. The real cost is everything the brand touches, so consider a refresh first and protect what people actually recognise. Most rebrands are triggered by internal boredom and justified afterwards with research. The few that are worth doing are triggered by a business fact that has already changed. A rebrand throws away recognition you have already paid for. Sometimes that is exactly right, because the recognition is for something you no longer are. Usually it is an expensive way of giving a new stakeholder something visible to point at. ## The four reasons that justify it - The business changed materially. A merger, a new category, a shift from one audience to another. The name or the position now describes something you do not do. - The brand actively blocks growth. It reads as regional when you sell nationally, as cheap when you sell premium, or as a single product when you sell a portfolio. - A legal or practical constraint. A trademark conflict, a domain you cannot get, a name that fails in a market you are entering. - Reputation damage severe enough that the name itself is the liability. Rare, and usually better addressed by fixing the cause first. Not on the list: the leadership team is tired of it, a competitor did one, the logo looks dated, or a new CMO has arrived. Those are the four most common actual triggers. ## What it really costs Design is the small part. The real cost is everything the brand touches: signage, packaging, uniforms, vehicles, templates, the product interface, email domains, legal filings, marketplace listings, partner collateral, and every third-party listing that references you. Add the internal time to run it, and the months where marketing effort goes into the transition rather than into demand. Then add the equity cost: search rankings, brand recall, existing bookmarks and links. That can be managed, but only deliberately: a name change is a migration, and migrations fail in the ways set out in redesigning without losing your rankings (https://colleviate.com/blog/website-redesign-without-losing-rankings.html). ## Refresh versus rebrand A refresh keeps the recognisable assets — the name, the core mark, the colour — and modernises the system around them. It costs a fraction, keeps the equity, and solves most of the problems people bring to a rebrand. If the honest answer to “what business fact has changed” is “none, it just looks old”, you want a refresh. ## Protect what people actually recognise Recognition attaches to specific assets, and they are rarely the ones the design review is about. A colour, a shape, a sound, a layout, a phrase. Find out which of yours are actually recognised before you change anything, then protect those through the transition even if everything around them moves. Throwing away a distinctive asset that took a decade to build, because it no longer suits the new typeface, is the most common self-inflicted cost in this whole exercise. The broader argument is in audacity and tenacity (https://colleviate.com/blog/audacity-and-tenacity.html). ## How to run the transition Announce it once, explain the reason in plain terms, and run both identities together for a defined period rather than switching overnight. Tell partners and marketplaces before customers notice. Keep the old domain redirecting permanently, not for a year. And resist relaunching everything else at the same time: a simultaneous rebrand, website rebuild and repositioning makes it impossible to tell which one worked. Related. If the name is the problem, see brand naming (https://colleviate.com/blog/brand-naming.html); for the launch that follows, a brand launch campaign plan (https://colleviate.com/blog/brand-launch-campaign-plan.html). ## Questions **How much does a rebrand cost?** Design is a minority of it. Budget for signage, packaging, templates, legal filings, listings, internal time and the migration itself, which together usually exceed the creative fee several times over. **Will a rebrand hurt our search rankings?** A name and domain change will, unless it is run as a migration with full redirect mapping and a deliberate plan. A visual refresh on the same domain generally will not. **How do we know if we need a rebrand or a refresh?** Ask which business fact has changed. If the honest answer is none, you want a refresh: it keeps the recognition you have already paid for. --- # Share of search: the brand metric that leads revenue URL: https://colleviate.com/blog/share-of-search-leading-indicator.html Topic: Share of search | Published: 2026-07-13 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Share of search is the proportion of branded search interest a company captures against its named competitors. It leads revenue because a branded search is intent forming before purchase; research popularized by Les Binet and the IPA found changes in share of search tend to precede changes in market share by several months. Most of the numbers a marketing team reports describe the past. Revenue, market share, and pipeline are all confirmations of demand that already converted. Share of search is one of the few brand metrics that moves before revenue does, and it can be read entirely from public signals, without a single line of first-party data. Share of search is the proportion of branded search interest a company captures against its named competitors in a category. If four brands together account for one hundred thousand branded queries in a period, and yours accounts for forty thousand of them, your share of search is forty percent. The inputs are branded search volumes; the output is a single position against a real competitive set. ## What it actually measures The distinction that matters is between interest and purchase. Market share counts transactions that already happened. Share of search counts the moment a person types a brand name into a search box, which is intent forming before any money changes hands. That is why the two metrics move at different times: one is a record of demand, the other is a leading trace of it. Because branded search is a deliberate act rather than a passive impression, it filters out much of the noise that makes upper-funnel awareness metrics hard to trust. Someone searching your name has already decided you are worth considering. Aggregate that decision across a category and you get a clean read on where mental availability is accumulating and where it is leaking. ## Why it leads revenue The reason share of search earns board attention is empirical, not theoretical. Marketing-effectiveness research popularized by Les Binet and the IPA found that a brand's share of search explains a large majority of its market share on average across categories, and that changes in share of search tend to precede changes in market share by several months. A rising share of search is one of the earliest observable signals that demand is shifting in your favor. A falling one is an early warning that a competitor is winning the consideration stage while your revenue line still looks healthy. > Revenue tells you what the market already decided. Share of search tells you what it is deciding now. The gap between the two is your window to act. That lead time is the entire value. By the time a market-share loss appears in the financials, the demand shift that caused it happened a quarter or two earlier. A brand watching only lagging numbers is always responding to a decision the market has finished making. A brand watching share of search is responding while the decision is still in motion. ## Reading it as an outside-in signal Share of search is a natural fit for outside-in analysis because every input is public. You do not need access to a company's analytics to compute it; you need its branded query volume and its competitors', both observable through search-trend data. That makes it as readable for a company assessing a rival, or an investor assessing a holding, as it is for the brand assessing itself. The measurement discipline is where most attempts go wrong. Share of search is only meaningful against the right competitive set, so the first step is defining the peer group honestly rather than flatteringly. It also has to be normalized for category seasonality and cleaned of ambiguous or homonym-heavy brand terms, or the reading drifts. Done carelessly, it produces a confident number that means nothing. Done with a defined peer set and consistent methodology, it becomes one of the cleaner leading indicators available. The same peer-set discipline underpins a full outside-in brand audit (https://colleviate.com/blog/outside-in-brand-audit.html). ## From metric to friction On its own, share of search answers one question: is consideration moving toward you or away from you. That is valuable, but it is a symptom, not a diagnosis. A falling share of search does not tell you whether the cause is a weakening product story, a competitor outspending you, a broken funnel that wastes the demand you do capture, or a search surface where answer engines have stopped citing you (https://colleviate.com/blog/how-ai-answer-engines-choose-sources.html). The signal points to where growth is leaking without explaining why. This is where a single metric becomes part of a larger read. Share of search sits alongside other public traces of demand and conversion, and each one converts into an estimated revenue effect. Together they add up to one evidenced reading of where a brand’s growth is constrained and by how much, scored against the brand’s real competitive set, so the diagnosis is grounded before we decide what to change. The point is not to admire the leading indicator, but to act on what it is warning you about while there is still time to change the outcome. ## What does the evidence for share of search actually say? The measure was set out by Les Binet at the IPA's EffWorks Global event in 2020. He tested it in three categories, cars, energy and mobile handsets, and found share of search tracked share of market in all three, with share of search moving first. The lead time differed by category: up to a year for cars, about six months for handsets, and from zero to three months for energy. Binet was also clear that it is not a perfect predictor, because conversion depends on other factors, price above all. The pattern behind those lead times is the length of the buying cycle. People research a car for months and switch an energy supplier in an afternoon, so the gap between searching and buying, and therefore the warning share of search gives you, stretches or shrinks with the category. ## What does it look like in a real market? We publish the Colleviate Share of Search Index (https://colleviate.com/blog/share-of-search-index-india.html) for nine Indian categories, built exactly as this article describes. It shows the kind of movement the measure exists to catch: in electric two-wheelers, Ather rose from 35.2 to 62.2 percent of the category's branded search between late 2021 and late 2026, while Ola Electric fell from 48.7 to 14.7 percent. In stockbroking, Groww has led Zerodha every month since September 2025. Those are readings of consideration, not sales. Whether market share follows is exactly the question the measure lets a board ask early. ## Sources - Binet presents fast, cheap, predictive Share of Search metric, IPA, 2020: https://ipa.co.uk/news/binet-presents-fast-cheap-predictive-share-of-search-metric/ - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What is share of search?** Share of search is the proportion of branded search interest a brand captures against its named competitors in a category. If four brands account for 100,000 branded searches and yours accounts for 40,000, your share of search is 40 percent. **Why does share of search lead market share?** A branded search is intent forming before any money changes hands, while market share records purchases that already happened. Research popularized by Les Binet and the IPA found that changes in share of search tend to precede changes in market share by several months. **Can you measure a competitor's share of search?** Yes. Every input is public, so share of search can be computed from search-trend data without access to anyone's analytics. It is only meaningful against an honestly defined competitive set, normalized for seasonality and cleaned of ambiguous brand terms. --- # The marketing measurement gap: why spend resists attribution URL: https://colleviate.com/blog/marketing-measurement-gap.html Topic: Measurement | Published: 2026-07-06 | Updated: 2026-10-14 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Marketing spend resists attribution because the gap is structural. Marketing and sales optimize different numbers, data is split across ad platforms, site, CRM and billing, and each platform credits itself, so attributed totals exceed real revenue. The gap narrows with an outside view of the whole buying surface, priced with evidence. Every CMO has faced the same question from the CFO: what did that budget actually return? The honest answer is uncomfortable, because most marketing spend cannot be cleanly traced to revenue, and the tools that claim to trace it are quietly scoring their own homework. The gap between marketing spend and provable revenue is not a sign of a lazy team or a bad dashboard. It is structural. In a 2024 Gartner survey of 378 senior marketing leaders, only 52 percent said they were succeeding at proving marketing’s value and receiving credit for it, and the reason is not communication. It is that the measurement itself is fragmented, self-interested, and blind to most of the journey. Closing that gap starts with understanding why it exists in the first place. ## Why the gap is structural, not fixable with more dashboards Three forces keep marketing spend from resolving into a clean revenue number, and none of them is solved by buying another analytics tool: - Split metrics: marketing is usually measured on lead volume while sales is measured on closed revenue. The two teams optimize different numbers, so the handoff where a lead becomes revenue is exactly where the tracking breaks. - Data silos: ad platforms, the site, the CRM, and the billing system each hold one fragment of the path. First touch is often lost, and no single system sees the whole journey from first impression to signed contract. - Self-interested attribution: each ad platform reports revenue by its own methodology, crediting itself for purchases it merely touched. When a buyer is influenced by several channels, every platform can claim the same sale, so the attributed totals add up to more than the revenue that actually happened. The result is a dashboard that looks precise and is not. It reports a number to two decimal places while missing most of what moved the buyer. That false precision is worse than honest uncertainty, because it invites decisions built on a measurement no one can audit. > A marketing dashboard that overstates its own contribution is not neutral. It is an argument disguised as a fact, and the CFO can usually tell. ## What the platforms cannot see Attribution tools watch clicks. They do not watch the parts of the buying decision that happen off-platform, and those parts are often decisive. A prospect who arrives ready to buy may have been moved by a slow, broken site that quietly cost the conversion, by a competitor's stronger position, or by public sentiment the campaign never touched. These forces shape revenue but never appear in a click path. A brand that measures only what its platforms report is measuring the narrowest slice of what actually determines whether spend converts. ## Closing the gap with the outside view The measurement gap narrows when the read moves from inside the ad account to the whole surface a buyer experiences. An outside-in brand audit (https://colleviate.com/blog/outside-in-brand-audit.html) looks at the full path to purchase the way the market does: the friction on the site, the gaps against competitors, the sentiment around the brand, and the alignment between the message and what buyers actually encounter. Much of what suppresses marketing ROI is not the campaign at all. It is technical debt on the conversion surface and other forms of friction that no attribution model was built to catch. ## From spend justification to a single friction reading The point is not to build one more attribution model to argue with the last one. It is to reframe the question. Instead of asking which channel deserves credit, ask where growth is actually being lost across the whole funnel, and price it with evidence. Each source of friction converts into an estimated revenue effect and adds up to one evidenced reading of where a brand is leaking growth, re-scored as the work runs, and the number arrives with its evidence rather than our opinion. That is a number a CMO can take into the room with the CFO, because it is grounded in evidence rather than in a platform's account of its own worth. ## Why can platform-reported results not close the gap? Because the methods platforms and most dashboards rely on are observational, and observational estimates of advertising effect are often wrong. The clearest evidence comes from Gordon and colleagues, published in Marketing Science in 2019. They compared fifteen large randomized advertising experiments at Facebook, covering about 500 million user-experiment observations and 1.6 billion ad impressions, with the observational methods commonly used in industry. The observational methods frequently failed to reproduce the experimental results, even after conditioning on extensive demographic and behavioral data. That is the structural point in one finding: a model built on who saw an ad and who bought cannot reliably separate the ad's effect from the fact that the people who see ads were already more likely to buy. ## What do the open-source models change? They make a second view cheap. In January 2025 Google made Meridian, its open-source marketing mix model, available to all marketers; it uses Bayesian methods to estimate media effects with stated uncertainty and can include pricing and promotions. A mix model does not close the gap on its own, because it is only as good as the baselines and experiments used to calibrate it. But it moves the question from which channel the platform credits to what the evidence supports. ## Sources - Gartner survey finds only 52% of senior marketing leaders can prove marketing’s value and receive credit, Gartner, via Business Wire, 18 September 2024: https://www.businesswire.com/news/home/20240918141146/en - A comparison of approaches to advertising measurement: evidence from big field experiments at Facebook, Marketing Science 38(2), 2019: https://pubsonline.informs.org/doi/10.1287/mksc.2018.1135 - Google opens Meridian, its marketing measurement tool, to all users, Search Engine Land, January 2025: https://searchengineland.com/google-meridian-all-users-451286 ## Questions **Why can't marketing spend be traced cleanly to revenue?** Because the gap is structural: marketing is measured on leads while sales is measured on revenue, the journey is split across ad platforms, site, CRM and billing, and each platform credits itself for sales it merely touched, so attributed totals add up to more than the revenue that happened. **Will a better attribution tool close the measurement gap?** Not on its own. Attribution tools watch clicks, and much of what decides a purchase happens off platform: site friction, a competitor's stronger position, public sentiment. Another model only re-argues credit for the narrow slice the platforms can see. **What should a CMO show the CFO instead?** A reading of where growth is being lost across the whole funnel, priced with evidence, rather than which channel deserves credit. A number grounded in evidence about the buying surface survives a finance conversation in a way a platform's account of its own worth does not. --- # The outside-in brand audit: see your company as the market does URL: https://colleviate.com/blog/outside-in-brand-audit.html Topic: Growth audits | Published: 2026-07-03 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** An outside-in brand audit measures the brand the market actually experiences rather than the one in its dashboards: digital presence, search visibility, customer sentiment, message coherence and competitive position. Benchmarked against the real peer set, it turns observations into decisions, and repeated regularly it becomes a continuous signal. Every brand has two versions of itself: the one in its dashboards and the one the market actually experiences. They diverge quietly, and the gap is where growth leaks. Internal analytics measure what a company chose to instrument, on surfaces it controls, interpreted by teams with an interest in the answer. Buyers experience something else: the page that loads slowly on a phone, the search result a competitor now owns, the review thread nobody in the building has read, the message that changed three rebrands ago and never fully propagated. An outside-in brand audit measures that second version, because it is the one revenue actually depends on. ## What an outside-in audit measures - Digital presence: real performance of the buying surfaces, broken paths in the routes buyers actually take, and the conversion drag technical debt creates. - Search visibility: which high-intent queries the brand owns, which it is ceding, and to whom, tracked as movement rather than a snapshot. - Customer sentiment: what reviews and public feedback say about the product, at the source, without a survey's framing effects. - Message coherence: whether the positioning survives the journey from ad to page to product, or fragments along the way. - Competitive position: all of the above, benchmarked against the real competitive set rather than the one in the brand deck. > The market audits your brand continuously whether you participate or not. The only choice is whether you see the results. ## Why benchmarks change the conversation A standalone audit produces observations. A benchmarked audit produces decisions. Knowing your load time is 6 seconds is trivia; knowing it sits in the worst quartile of your peer cluster, and roughly what that costs in conversion, is an agenda item. Benchmarking against the brand’s real competitive set is what turns brand health from a feeling into a position. ## From audit to a living signal The traditional brand audit is a project: commissioned, delivered, shelved. The signals it reads move weekly. Treated as a continuous reading instead, the audit is re-run on a fixed cadence through the engagement: when search position shifts, sentiment turns, or a competitor repositions, the change surfaces with evidence attached. For a CMO or CEO, that is the difference between an annual health check and a monitor. Related. The same habit applied to old advertising: how to read a vintage ad in five questions (https://colleviate.com/blog/how-to-read-a-vintage-ad.html). ## Questions **What does an outside-in brand audit measure?** The brand as buyers experience it: the real performance of the buying surfaces, search visibility on high-intent queries, customer sentiment at the source, whether the message holds from ad to page to product, and competitive position against the real peer set. **How is it different from an internal analytics review?** Internal analytics measure what a company chose to instrument, on surfaces it controls, interpreted by teams with an interest in the answer. An outside-in audit reads the public signals revenue actually depends on, including the parts no internal dashboard sees. **Why benchmark the audit against peers?** Because a standalone audit produces observations and a benchmarked one produces decisions. A slow load time is trivia until you know it sits in the worst quartile of your peer cluster and roughly what that costs in conversion. --- # AEO, GEO, and AI Visibility: How to Get a Company Cited by AI Answer Engines URL: https://colleviate.com/blog/aeo-geo-ai-visibility.html Topic: AI visibility | Published: 2026-07-03 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** Answer engine optimization (AEO) structures content so an AI system can find a specific answer, verify it and cite the page. Generative engine optimization (GEO) is the wider discipline of shaping how language models retrieve, synthesize and attribute sources. In practice both mean self-contained answers under question-led headings, schema that removes ambiguity, and claims a model can verify. Answer engine optimization (AEO) is the practice of structuring content so AI answer engines like ChatGPT and Perplexity extract, cite, and reuse it directly in generated answers. Generative engine optimization (GEO) extends this to influencing how large language models synthesize and attribute your content across every query where you are a relevant source. ## What is answer engine optimization (AEO)? Answer engine optimization is the practice of structuring content so an AI system can locate a specific answer, verify it, and cite the page as a source. It differs from ranking a link on a results page. The unit of success is a sentence or paragraph that a model can lift and attribute, not a blue link a user clicks. In practice, AEO means writing self-contained answers, marking them up with machine-readable structure, and making the underlying claims verifiable. Engines favor sources they can parse without ambiguity and trust without a second query. ## What is generative engine optimization (GEO)? Generative engine optimization is the practice of shaping how large language models synthesize and attribute content when they generate an answer rather than return a list. Where AEO focuses on the extractable answer unit, GEO addresses the broader synthesis: which sources a model draws from, how it paraphrases them, and which one it names. GEO and AEO overlap heavily and are often used interchangeably. The useful distinction is scope. AEO optimizes the answer on your page; GEO optimizes your presence across the model's full reasoning process, including retrieval, ranking, and attribution. ## How is optimizing for AI answers different from traditional SEO? Traditional SEO optimizes for a ranked list of links a human scans and clicks. AI answer optimization targets a single synthesized response where citation, not position, is the win. The click may never happen, so the goal shifts from earning a visit to earning an attribution. Three differences matter most. First, extractability replaces keyword density: the model needs a clean, quotable claim. Second, verifiability replaces link volume: a claim the model can confirm against structured data or a primary source is safer to cite. Third, the surface is conversational, so your content competes across many phrasings of the same question rather than one exact keyword. Understanding how AI answer engines choose sources (https://colleviate.com/blog/how-ai-answer-engines-choose-sources.html) is the foundation for all three. ## How do AI answer engines decide which sources to cite? AI answer engines cite sources they can retrieve, parse, and trust. Retrieval depends on the content being indexed and topically relevant to the query. Parsing depends on clear structure and a self-contained claim. Trust depends on signals like corroboration across sources, primary evidence, author and publisher identity, and freshness. Models are more likely to attribute a specific page when its claim is unambiguous, matches the query intent closely, and is backed by evidence the engine can corroborate. Content grounded in verifiable data with clear provenance is easier to cite safely, which is why grounded reasoning backed by evidence outperforms unsupported assertion. > An AI answer engine does not reward the page that argues hardest. It rewards the page that lets the model confirm a claim fastest and attribute it with the least risk. ## Which schema.org types matter most for AI answer engines? The schema.org types that matter most for AI answer engines are the ones that map content to a question or a fact: FAQPage, QAPage, HowTo, Article, and Organization. FAQPage and QAPage pair a question with a direct answer, which mirrors how engines retrieve. HowTo structures sequential steps. Article and Organization establish authorship, publisher, and entity identity. Supporting types add trust and disambiguation: Person and sameAs (https://colleviate.com/blog/organization-schema-and-sameas-for-agencies.html) connect an author to a verifiable identity, BreadcrumbList clarifies site structure, and Dataset signals structured evidence. Schema does not force a citation, but it removes ambiguity about what your content asserts and who stands behind it. ## How should you structure a page so an LLM can extract the answer? Structure a page so the answer to each question sits in the first two or three sentences under a heading that states the question. This inverted-pyramid pattern lets a model retrieve a heading, read the answer immediately below it, and quote it without parsing the whole document. Lead with the conclusion, then support it. Phrase headings as the literal questions people ask, keep each answer self-contained so it stands alone when lifted out of context, and place definitions and key facts in plain sentences rather than burying them in tables or images. Use consistent terminology for the same entity throughout, and make every factual claim traceable to a source the engine can check. ## A practical AEO and GEO checklist you can implement today The fastest wins come from making answers extractable and claims verifiable. Work through the list below on your highest-intent pages first, then extend it across the site. - Phrase every H2 as a question a person actually searches, and answer it in the first two or three sentences below. - Write each answer to be self-contained, so it holds meaning when a model lifts it out of the page. - Add FAQPage, QAPage, or HowTo schema where the content maps to questions or steps, plus Article and Organization for identity. - Back every factual claim with a verifiable source or dataset, and state numbers with their provenance. - Use one consistent name for each entity, product, or concept across the whole page. - Keep primary facts in plain text, not locked inside images, PDFs, or script-rendered elements a crawler may miss. - Establish author and publisher identity with Person, Organization, and sameAs links to verifiable profiles. - Refresh dated claims and timestamps so freshness signals stay current. - Cross-link related answers so engines can follow your coverage of a topic, as this Insights section does. AEO and GEO reward the same discipline we hold our own work to: a clear claim, a verifiable source, and unambiguous structure. Every Colleviate audit is built that way, with each figure carrying its source and its evidence tier, because a claim a client cannot check is a claim an answer engine will not cite either. We measure AI-answer citation share (https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html) for our clients against their named competitors as part of the Growth Audit (https://colleviate.com/blog/growth-audit-what-it-should-contain.html), and the programme to earn those citations runs inside Performance Marketing. ## How big is the AI answer surface now? Large enough that it is no longer a side channel. In July 2025 Alphabet reported that AI Overviews in Google Search had passed 2 billion monthly users across more than 200 countries and territories and 40 languages. Usage at that scale changes what a search result is: for many questions the first thing a person reads is a synthesized answer, and the links come after it. The effect on clicks has been measured. A Pew Research Center analysis of the browsing of 900 U.S. adults in March 2025 found that people clicked a traditional search result in 8 percent of visits when an AI summary appeared, against 15 percent when none did. Links inside the summary itself were clicked in 1 percent of visits, and people ended their browsing session on 26 percent of pages with a summary, against 16 percent without. Google disputed the methodology, but the direction is consistent with what publishers report: being cited inside the answer now matters as much as ranking beneath it. ## What does the research say actually works? The paper that named the discipline, GEO: Generative Engine Optimization by Aggarwal and colleagues, was accepted at KDD 2024. Testing content changes against a benchmark of queries across many domains, the authors report that GEO methods can boost a source's visibility by up to 40 percent in generative engine responses. Among the changes they tested, adding statistics and quotations from relevant sources were among the most effective, while keyword stuffing performed worst. They also note that "the efficacy of these strategies varies across domains," and the results come from a controlled benchmark rather than live traffic. The practical reading is consistent with everything else in this guide: sourced figures, quotable sentences and verifiable claims are what engines lift. The most durable version of that is publishing data nobody else has, which is why we publish our own share of search index for India (https://colleviate.com/blog/share-of-search-index-india.html) with its method and a free download. Next steps. To measure whether it is working, see how to track whether AI answers name your brand (https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html); if you are missing from answers, six checks, in order (https://colleviate.com/blog/why-ai-answers-dont-mention-your-brand.html); and on the new file everyone asks about, llms.txt, what it does and what to do first (https://colleviate.com/blog/llms-txt-what-it-does-and-what-to-do-first.html). Related. What works and what does not across AI tools: AI in marketing, what works (https://colleviate.com/blog/ai-in-marketing-what-works.html). ## Sources - Alphabet Q2 earnings call: CEO’s remarks, Google, July 2025: https://blog.google/company-news/inside-google/message-ceo/alphabet-earnings-q2-2025/ - Google users are less likely to click on links when an AI summary appears in the results, Pew Research Center, 22 July 2025: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/ - GEO: Generative Engine Optimization (Aggarwal et al., KDD 2024), arXiv, June 2024: https://arxiv.org/abs/2311.09735 - Colleviate Share of Search Index, India, Colleviate, October 2026: https://colleviate.com/blog/share-of-search-index-india.html ## Questions **What is the difference between AEO and GEO?** Answer engine optimization (AEO) optimizes the answer on your page: a self-contained, verifiable passage an AI system can lift and cite. Generative engine optimization (GEO) covers your presence across the model's whole process, including which sources it retrieves, how it synthesizes them and which one it names. The two overlap heavily; the useful distinction is scope. **How is optimizing for AI answers different from SEO?** SEO competes for position in a ranked list of links; AI answer optimization competes for citation in a single synthesized answer, where the click may never happen. Extractability replaces keyword density, verifiability replaces link volume, and content has to answer many phrasings of the same question rather than one exact keyword. **Which schema types help most with AI answer engines?** The types that map content to a question or a fact: FAQPage, QAPage, HowTo, Article and Organization. Supporting types such as BreadcrumbList, sameAs links and Dataset add disambiguation and trust. Schema does not force a citation, but it removes ambiguity about what a page asserts and who stands behind it. --- # How AI Answer Engines Decide Which Sources to Cite (and How to Become One) URL: https://colleviate.com/blog/how-ai-answer-engines-choose-sources.html Topic: AI visibility | Published: 2026-07-03 | Updated: 2026-10-29 Publisher: Colleviate (https://colleviate.com/) **Short answer.** AI answer engines retrieve a candidate set of sources from a search index and their own knowledge, then re-rank them on relevance, extractability, entity clarity and corroboration. Pages that answer the question directly, state claims plainly, name their data sources and match a verifiable entity get cited. Being indexed is not enough; being cleanly quotable is. AI answer engines choose sources by retrieving a candidate set from a search index and internal knowledge, then re-ranking those candidates on relevance, extractability, entity clarity, and corroboration. Pages that state claims plainly, name their data sources, and match a verifiable entity get cited most. Being indexed is not enough. Being cleanly liftable is what earns the citation. ## How does an AI answer engine retrieve and rank candidate sources? An answer engine first retrieves a broad candidate set using a search index, embedding similarity, or both, then re-ranks that set against the specific question. Retrieval decides which pages are eligible. Re-ranking decides which few actually get read into the answer. Most sources that are indexed never make it past re-ranking because they do not answer the question directly enough. The practical implication is that ranking for a keyword and being cited for a question are different problems. The engine scores each candidate on semantic closeness to the query intent, then filters for passages it can quote with confidence. For the wider mechanics of this shift, see the AEO and GEO field guide (https://colleviate.com/blog/aeo-geo-ai-visibility.html). ## Why does extractability decide whether your page gets quoted? Extractability decides quoting because an engine can only cite a claim it can isolate as a clean, self-contained statement. If the answer to a question is buried across three paragraphs, or depends on a chart with no text equivalent, the model cannot lift it reliably and moves to a page that states it plainly. A direct sentence beats a well-written page that circles the point. Structure this deliberately. Put the answer in the first two sentences under a heading phrased as the question, keep claims atomic, and give every figure a written form next to any visual. This is inverted-pyramid writing applied to machine reading. ## What does an engine verify about entity identity before it cites you? Before citing you, an engine tries to resolve who you are and whether the entity making the claim is consistent and real. It cross-references your organization name, domain, author, and described identity against other mentions of the same entity. Ambiguity or contradiction between your site and the wider record lowers the confidence needed to quote you. This is where a knowledge graph matters. Consistent entity signals across your site, structured data, and third-party references let the engine bind a claim to a stable identity. Reading a brand from public signals to build that cross-source picture is exactly what an outside-in brand audit (https://colleviate.com/blog/outside-in-brand-audit.html) does, and it is the same logic an engine applies in reverse when it decides whether to trust you. ## Why do cited statistics with named sources get lifted more often? A statistic with a named source and a date gets lifted more often because it carries its own verification. When a page states a figure, attributes it, and dates it, the engine can present that claim with a provenance chain instead of an unsupported assertion. Naked numbers are riskier to quote and get passed over. > An engine is not looking for the best-written page. It is looking for the claim it can defend when a user asks where the answer came from. Grounding a claim in a citation is the difference between a quotable fact and an opinion the model must hedge. It is the same rule we apply to our own audits: a figure without a source is not a finding. Engines prefer corroborated claims for the same reason clients do. ## How do E-E-A-T and author identity affect whether you are cited? Experience, expertise, authoritativeness, and trustworthiness affect citation because engines weight sources by how much the wider record supports the author and publisher. A named author with a verifiable track record, a domain with topical depth, and consistent expertise signals raise the odds that a passage survives re-ranking. Anonymous or thin pages are cited only when nothing better exists. Author identity works the same way entity identity does. Attach real authors, describe their relevant experience, and let those signals be corroborated elsewhere. Trust is inferred from the record, not asserted on the page. ## Where do AI answers actually pull from besides your own site? AI answers pull from far more than your site, including reference encyclopedias, structured databases, review platforms, community forums, industry publications, and any high-authority page that discusses your entity. The engine assembles an answer from the strongest available sources, so your own claims compete with, and are checked against, everyone else's. This means off-site corroboration is part of your citation strategy. If independent sources describe your entity consistently, the engine gains confidence to cite your own statements. If the wider record is silent or contradictory, your site alone rarely carries the answer. ## How do you become a cited source? A technical checklist You become a cited source by making claims easy to retrieve, easy to extract, and easy to verify. Work through the following. - Phrase headings as the literal questions users ask, and answer each in the first two or three sentences. - State one claim per passage in a self-contained sentence that reads correctly out of context. - Attribute every statistic to a named source with a date, and give visuals a written equivalent. - Add structured data for organization, author, and article so entity identity resolves cleanly. - Attach named authors with verifiable, described expertise on the topic. - Keep entity naming consistent across your site, metadata, and external profiles. - Earn corroborating mentions on reference, community, and industry sources the engine already trusts. - Maintain a crawlable, fast, well-linked site so eligible pages are actually retrieved. ## How do you measure whether AI engines are citing your brand? You measure AI citation by querying the engines directly with the questions your audience asks and recording whether your brand appears, in what context, and with what accuracy. Track presence, sentiment, and factual correctness across engines over time, because a citation that misstates your position is a problem to fix, not a win. This is where AI citation stops being a guess and becomes a monitored surface. We baseline it for every client at the start of an engagement: the questions their buyers ask, which engines cite whom, and how that compares with named competitors. The checklist above then becomes a programme with a number attached, re-scored monthly alongside share of search (https://colleviate.com/blog/share-of-search-leading-indicator.html). ## What happens to clicks when an answer appears? They fall, which is why citation matters more than position. Pew Research Center's analysis of 900 U.S. adults' browsing in March 2025 found traditional result links were clicked in 8 percent of visits when an AI summary appeared, against 15 percent without one, and links inside the summary were clicked in just 1 percent of visits. When the answer is read and not clicked, the only visibility left is being named in it. Google disputed the study's methodology; the measured direction is still the one to plan for. ## Which content changes measurably raise citation? The most rigorous public evidence is the GEO study by Aggarwal and colleagues, accepted at KDD 2024, which reports visibility gains of up to 40 percent in generative engine responses from content-level changes. Adding statistics and adding quotations from relevant sources were among the strongest methods it tested, and keyword stuffing among the weakest. The authors caution that effectiveness varies by domain, and the experiments ran on a benchmark, not on live engines at scale. Both findings point the same way as the checklist above: a claim that arrives with its number, its source and its date is easier for an engine to defend, and therefore easier to cite. Related. Why AI answers do not mention your brand (https://colleviate.com/blog/why-ai-answers-dont-mention-your-brand.html) turns this into a checklist, and how to track whether AI answers name your brand (https://colleviate.com/blog/how-to-track-ai-citations-for-your-brand.html) shows how to measure the result. ## Sources - Google users are less likely to click on links when an AI summary appears in the results, Pew Research Center, 22 July 2025: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/ - GEO: Generative Engine Optimization (Aggarwal et al., KDD 2024), arXiv, June 2024: https://arxiv.org/abs/2311.09735 ## Questions **Does ranking well in Google mean AI engines will cite you?** Not necessarily. Ranking for a keyword and being cited for a question are different problems. Retrieval decides which pages are eligible; re-ranking decides which few are read into the answer, and most indexed pages fail there because they do not answer the question directly enough. **Why do statistics with named sources get cited more often?** A figure that is attributed to a named source and dated carries its own verification, so an engine can present it with a provenance chain. Unsourced numbers are riskier to quote and tend to be passed over in favor of a claim the engine can defend. **How can you tell whether AI engines are citing your brand?** Ask the engines the questions your buyers ask and record whether your brand appears, in what context and how accurately. Track presence, sentiment and factual correctness across engines over time, and compare against named competitors, because a citation that misstates your position is a problem rather than a win.