Earned media has a second job now

For most of its life, earned coverage sold one thing: credibility. Now it also feeds the machine that answers your customers’ questions. The brands getting cited are the ones a real outlet vouched for first.

For most of its life, earned media did one thing well. A credible outlet wrote about you, and their credibility rubbed off. A reader trusted the magazine, so they extended a little of that trust to the brand inside it. That was the whole transaction. Third-party validation, the textbook called it.

That job hasn't gone away. But there's a second one now, and most brands haven't noticed it yet.

When someone asks an AI assistant for the best option in your category, the answer doesn't come from your website. It comes from what the rest of the internet has said about you, weighted by how much the system trusts the source. A model assembling an answer is reading the same coverage a customer used to read. It is deciding, in effect, who counts. And the brands it names are, more often than not, the ones a real publication wrote about first.

So the press release you placed last spring is no longer just sitting in an archive. It is training data. It is the raw material an answer engine pulls from when it decides whether you exist.

The industry is starting to say this out loud. At this year's Meltwater Summit, Dara Busch, CEO of Havas PR, said earned media is "having a glow up." Her reason was specific: the way earned coverage now powers AI. She also named a shift that's been a long time coming. For decades, PR was sold on circulation and impressions, on the size of the audience. That's over. The question now, she said, is who actually saw it, and whether the coverage connects to something that moved.

Here is the part worth sitting with, because it cuts against the obvious response.

The obvious response is volume. If coverage feeds the machine, get more coverage. Flood every outlet, chase every mention, treat earned media like a numbers game. That instinct is wrong, and it's wrong in a way that matters.

An answer engine isn't counting your mentions. It's weighing them. A hundred placements in outlets nobody trusts teach it nothing, except possibly that you're trying too hard. One placement in a publication it treats as authoritative does more work than the hundred combined. The whole system runs on the editorial judgment of the outlet that ran the story. Which means the value was never in the volume. It was in the quality of the source and the quality of the story that it earned its way in.

That is a very old idea wearing new clothes. The editors who decide what's worth covering are now, indirectly, deciding what the machine repeats. Getting a real story into a real outlet was always the hard part and the valuable part. It's just that the payoff used to end with the reader. Now it carries forward into every AI answer that outlet helps shape.

None of this makes the work easier. It makes it more legible. The brands that win the next few years won't be the loudest. They'll be the ones with something specific enough to be worth reporting, placed somewhere credible enough to be worth citing. That has always been the job. The stakes just got higher because the coverage keeps answering questions long after the reader has moved on.

There's a line on the Wink homepage that has started to read differently: there's nothing artificial about our intelligence. It was a position about how we work. It's becoming a description of the whole problem. The intelligence answering your customer's question is artificial. What it has to go on shouldn't be. Earned media is how you put something real and reported into a system that is otherwise guessing.

That's the job. It hasn't changed. It just started to matter in a second place.