Customer research when you cannot afford research
The expensive version is a survey of a thousand people answering questions they have not thought about. The cheap version is better.
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.
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.
Questions we get asked
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.
Start with a Growth Audit
Every Colleviate engagement begins with an outside-in read of your brand and its competitive cluster, before we recommend a single line of spend.
Ask for a scoped auditWhat we do · How we produce it · Who we do it for · Talk to us