PR worth paying for

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.

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.