The 'Earned Media' Debate Is a Distraction From What Matters

There's a very PR thing happening right now: a genuinely useful data set dropped, and instead of figuring out what to do with it, people started arguing about definitions. Classic.

The debate kicked off after Muck Rack published their "What Is AI Reading?" research — three editions now, spanning July 2025 through May 2026 — which analyzed more than 25 million links cited by ChatGPT, Claude, and Gemini. The finding that got people talking: roughly 82–89% of AI citations come from non-owned, non-paid sources. Muck Rack's VPs of Data & Intelligence and Communications, Linda Zebian and Linda Zebian, called that category "earned media."

Cue the discourse.

Some practitioners pushed back on the label. Should Wikipedia really count as earned? What about Reddit threads and government sites? Is this inflating the role of PR?

These are fair questions — but they're also, frankly, a little beside the point.

THE DEFINITION DEBATE IS AS OLD AS THE INDUSTRY

"Earned" has never had a clean, universal definition. Some PR pros use it to mean press coverage and only press coverage. Others use it to mean anything a brand didn't buy or publish itself. According to Dzugan and Zebian, their broader definition reflects how much the PR job has evolved — their 2026 State of PR survey of more than 1,000 PR professionals found that content creation and influencer work now rank among the top job functions, and most teams are running LinkedIn and Instagram as core communications channels.

So yes, the bucket is wide. And they publish the individual category breakdowns every time precisely so you don't have to take their word for how the sausage gets made.

The journalism number specifically? Remarkably steady. Journalism has accounted for 25–27% of AI citations across all three research editions. That's one in four citations, consistently, across three of the biggest AI platforms on the internet. That's not nothing.

THE WIKIPEDIA QUESTION IS MORE INTERESTING THAN IT SOUNDS

Wikipedia is the largest non-journalism source in Muck Rack's data and the most-cited domain on ChatGPT in their research. Critics say lumping it in with earned media is a stretch.

But here's the part worth sitting with: Muck Rack analyzed 489 S&P 500 company Wikipedia pages and found that 60% of the 58,490 references cited across them came from journalism. Wikipedia requires independent, reliable sourcing to justify a company's page existing at all. So when AI cites Wikipedia, it's often citing a downstream artifact of press coverage.

That doesn't make Wikipedia journalism. But it does mean you can't fully separate your Wikipedia presence from your media relations program. They're linked — often directly.

WHAT THIS MEANS FOR YOUR PR STRATEGY

A few things to take away from all of this, beyond the definitional drama:

• Generative engine optimization (GEO) is not a media relations replacement — it's an expansion. Media relations alone doesn't explain 84% of AI visibility. There are too many other source types involved. But strong press coverage still feeds the broader ecosystem in ways that matter.
• Think about the citation chain, not just the final URL. A news article might get cited in a Wikipedia page, discussed on Reddit, referenced on a trade site — and eventually show up in an AI answer once removed from where it started. That upstream influence is hard to measure but very real.
• Your Wikipedia presence is now a GEO priority. If your company has a sparse or unsourced Wikipedia page, that's not a vanity problem anymore. It's a visibility problem in AI platforms.
• Don't expect one number to tell the whole story. AirOps analyzed 21,311 brand mentions and found 85% came from third-party sources — different methodology, similar directional finding. The pattern is consistent: AI leans heavily on what it didn't get from you directly.

The big shift here is less about redefining earned media and more about recognizing that every credible third-party signal you generate — press hits, Wikipedia citations, academic references, online community discussions — is now potentially training material for the answers AI gives about your brand. That should change how you think about the value of long-term reputation work, not just the campaign you're running this quarter.

Stop arguing about vocabulary. Start auditing your AI footprint.

Want help thinking through what this means for your communications strategy? Winsome Marketing works with brands navigating exactly this kind of shift — where traditional PR meets the new reality of AI-driven visibility. Let's talk.

This post was originally inspired by What does 'earned' mean in the age of AI? via prdaily. We encourage you to read the original piece for full context.