TL;DR

A client recently begged me not to send him another article about AI visibility—and he had a point. But the latest one worth your time argues that AI systems reward probability, not popularity: LLMs cross-reference your owned media, earned coverage, and community discussion to decide whether your claims are credible. 

That’s not a new discipline called “unpaid media.” It’s the PESO Model®, described in near-perfect detail without ever being named—which means if you’ve been doing integrated communications, you’ve been building AI visibility all along. Same strategy. Different tools.

Key Insights:

• “Unpaid media” is owned, earned, and shared media. When an article prescribes a convergence of those things, it’s describing the PESO Model®—it just isn’t naming it. (It’s called communications, my friends.)
• AI engines are three times more likely to cite premium publisher content than brand-owned content, which turns media relations from a warm-feelings awareness play into a discoverability engine.
• The traffic that does arrive from AI referrals converts at up to four times the rate of traditional organic search—fewer people are kicking the tires, but the ones who do are far more likely to buy. Don’t forsake conversions for traffic.
• Every PESO media type has one job in the evidence base: owned media states the claim, earned media validates it, shared media generates the community signals, and paid media amplifies reach—but paid cannot build credibility on its own.
• Inconsistency across channels doesn’t just dilute your message anymore—it lowers your confidence score. The cross-team coordination the industry is now calling for is the integration problem the PESO Model was built to solve more than a decade ago.

AI Visibility is a PESO Model® Problem—and You’re Closer Than You Think

A few weeks ago, I told a client I was sending over an article I thought his team should read.

“Let me guess,” he said. “It’s about AI visibility.”

It was, in fact, about AI visibility.

“Great. Exactly what I need. Another article about AI visibility.”

Fair.

(And if you’re reading this, you know who you are. You were right! Go ahead and print this and use it against me later, YOU WERE RIGHT!)

He was right because the AI visibility content machine has become exhausting. Every consultancy, every martech vendor, every LinkedIn thought leader has published some version of the same piece: AI changed search, you should be panicking, and—what luck!—they happen to sell the thing that will fix it. 

The framing is always the same, too. 

AI visibility gets presented as a brand-new discipline requiring brand-new experts, brand-new tools, and a brand-new budget line.

Nope.

Nothing has changed strategically.

Same strategy. Different tools.

And that is the genesis for our conversation today. Not to add to the pile, but to pull one article out of it. A piece in Marketing Dive called Why Unpaid Media Is Now Essential to AI Visibility, and it’s worth your time for one very specific reason.

It describes the PESO Model® in near-perfect detail.

It just never names it.

But also, unpaid media? It’s called communications, my friends. Not unpaid media. As if.

The Article that Describes the PESO Model Without Naming it

Let’s look at what the Marketing Dive article actually says, beyond the dumb headline. 

AI systems cross-reference owned media, earned media, third-party validation, and community discussion (aka shared media) to determine how much confidence to place in a brand’s self-claims. 

Brands that show up in AI-generated answers are the ones with a consistent, verifiable presence across the sources those systems trust. 

The author’s prescription? 

This new hip thing called “unpaid media”—her term for earned media coverage and community discussion—working in concert with owned media and paid amplification. 

A convergence, as she puts it, of paid, owned, and earned media.

Gosh. If only someone had come up with this sooner.

To be clear, this isn’t a dunk on the author. Her analysis is genuinely good. 

When a smart strategist sits down to figure out how AI visibility actually works and independently arrives at an integrated model of paid, earned, shared, and owned media, that’s not a new idea or plagiarism. It’s convergent evolution. The framework won the argument.

And, you’re already ahead of the game if you’re using PESO. 

If you’ve been running an integrated PESO Model program, you haven’t been waiting for the AI visibility era to arrive.

You’ve been preparing for it. Possibly for years. You just didn’t know this was the payoff. 

(Neither did I, to be fair.)

AI Systems Reward Probability, Not Popularity

The single best line in the article—the one I’d tattoo on the arm of everyone still reporting impressions—is this, “AI systems don’t reward popularity, they reward probability, specifically, how confident the model is that what a brand claims is reputable.”

And that, my friends, is the measurement shift we’ve been talking about all year. It’s not “unpaid media.” It’s what you do every single day. It’s how you prove that the claims you make are true via third-party sources. 

Popularity is what our vanity metrics measure. Impressions, reach, AVEs, follower counts—they all answer the same question: how many eyeballs passed by? And for two decades, we’ve let those numbers stand in for effectiveness because they were easy to get and because they show an upward arrow to the right on a graph.

Probability is a different question entirely. 

When someone asks an AI to recommend a brand in an industry they’re interested in buying from, the LLMs aren’t checking your follower count or how many impressions you garnered last quarter. 

They’re running something closer to a credibility audit to figure out whether the brand’s website reflects what the media coverage and online communities say. 

It wants to know if the Reddit thread agrees with the case study (though Reddit and OpenAI are having a tiff, so that may not be the case for much longer). 

It wants to know if the analyst mentions corroborate the claims made on the About page.

Every consistent answer raises the model’s confidence. Every gap lowers it.

The wrong conversation is “how do we get more reach?” 

The right question is, “How much of what we claim about ourselves can the internet independently confirm?”

That second question should make some of us squirm a little. It’s supposed to. AND. AND! It’s what we do. It’s what we’ve always done.

The strategy hasn’t changed. The tools to do the job have changed. 

Every PESO Media Type Has a Job in the Evidence Base

Here’s where the PESO Model stops being a nice organizing framework and starts being an operating system. 

Each media type contributes something specific to the body of evidence that AI systems evaluate—and no media type can do another’s job.

Owned Media States the Claim

Your owned media is where the claim gets made. Your site, your blog, your case studies—this is the base layer, and it needs to be structured so machines can actually parse it. 

It should have clear answers to real questions, consistent messaging, and no burying your expertise in gated PDFs. 

But owned media is also, by definition, something you control, so, on its own, it’s not totally believable. That’s why you need validation and third-party credibility.

Enter earned media.

Earned Media Validates It

Now, we’ve come to realize over the past 18 months or so that AI engines are three times more likely to cite premium publisher content than brand-owned content.

Three times!

For years, media relations (aka “unpaid media”—it’ll never get old!) has been treated as the awareness play—nice coverage, warm feelings, impossible to measure. 

That era is over. 

When a journalist covers your company, a trade publication quotes your executive, your experts are interviewed on a podcast, a newsletter author features them, or an industry analyst cites your data, you’re not just reaching their audiences anymore. 

You’re creating the third-party validation that AI systems weigh most heavily when they decide whether to recommend you.

Media relations is now a discoverability engine. 

Shout it from the rooftops. Media relations is now a discoverability engine!
Our time has arrived.

Shared Media Generates the Community Signals

The Marketing Dive article makes a point most AI visibility content skips: LLMs pay attention to community discussion. 

Reddit threads, LinkedIn conversations, niche forums, YouTube comments—the places where real people talk about your brand when you’re not in the room. 

These signals are hard to fake, which is precisely why the models trust them.

You can’t buy this kind of corroboration. You also can’t earn this kind of corroboration. It happens only when people buy from you and experience what it’s like to be a customer.

Paid Media Amplifies but Cannot Build Credibility

And then there’s paid. Ads now compete with an AI answer layer that resolves the question before the ad is ever seen. 

Paid still matters—it extends reach, it gets your best content in front of the right audiences faster. But it cannot manufacture credibility. 

A brand with great ads and no corroborating evidence is a well-lit empty store.

When all four work together, you’re not running campaigns anymore. You’re building a verifiable body of evidence—one that AI systems can check, confirm, and cite.

It’s a beautiful thing to behold when it all works together like that!

The Numbers That Should Change How You Measure

If you need help making this case internally, the data does most of the work.

Thirty-seven percent of consumers now start their searches with AI tools rather than Google. 

Sixty percent of searches end without a single click because the answer is delivered, and no website visit occurs. 

When Google’s AI Overviews appear, click-through rates for top organic listings drop by roughly a third. 

It’s a sad state of affairs to see our website traffic drop like that.

BUT! When someone visits your site from AI referrals, they convert at up to four times the rate of traditional organic search. 

So, while fewer people are visiting your site and kicking the tires, the ones who do are four times more likely to convert.

This reminds me of when I started my business. I would have coffees, lunches, drinks, and dinners with anyone and everyone. My calendar was stuffed full of meetings, but I wasn’t growing the business.

A coach I worked with at the time suggested I stop saying yes to all of the meetings and focus only on the ones that would drive new business for us.

It was terrifying! My calendar suddenly went from 10 meetings a day to three.

But guess what happened? 

The meetings I had were valuable, and I was able to grow the business.

Don’t forsake conversions for traffic. 

I digress, but this means measurement has to evolve past clicks and conversions. 

We have to ask ourselves: 

• Are we included in AI Overviews when they summarize our category? 
• How often do LLMs mention us? 
• What’s our share of voice in AI-generated answers versus our competitors’? 

Those metrics feel unfamiliar today. So did domain authority and backlinks once upon a time.

The Integration Problem We Already Solved

Near the end of the Marketing Dive article, it details that all of this requires “closer coordination across paid media, SEO, content, PR, social, and analytics than most organizations are used to.”

Than most organizations are used to. That’s doing a lot of work in that sentence, isn’t it?

Because we all know what this looks like in practice: the SEO team optimizing for keywords nobody on the PR team knows about, the paid team bidding on terms the content team never wrote for, the social team off running engagement plays nobody can connect to revenue. 

Every team hitting its own metrics. No team building a coherent evidence base for humans AND LLMs..

AI systems are ruthless about this fragmentation because inconsistency across channels doesn’t just dilute your message, it lowers your confidence score.

This is the integration problem the PESO Model was built to solve—not last quarter, but more than a decade ago. 

One strategy, four media types, each doing its distinct job in service of the same verifiable story. The pros who’ve been running it this way don’t need an AI visibility wake-up call.

They are already awake.

It’s not unpaid media. It’s the PESO Model, driven by owned, earned, and shared media.

What to Do With Your Head Start

So if you’re a PESO Model practitioner, here’s your homework—and it’s more satisfying than most homework, because it’s work you’ve already done. Which I guess isn’t technically homework. 

Start with this question from the Marketing Dive article, “What does the internet reliably confirm about us, and is that enough for AI to recommend us?”

Then map your claims against your corroboration. 

Take the three most important things your brand says about itself—your positioning, your differentiators, and the expertise you want to be known for. 

For each one, ask: 

• Where does earned media confirm this? 
• Where is the community discussing it? 

If a claim lives only on your own website, it isn’t evidence yet. It’s an assertion waiting for validation—and now you know exactly which earned and shared plays to run.

And finally, take this to your leadership team, not as another AI panic memo, but as the opposite. The message isn’t “we’re behind and need a new budget line.” 

The message is “the integrated program we’ve been running is now the thing that determines whether AI recommends us, and here’s where we double down.”

That’s a much better meeting (and you might finally, FINALLY be able to stop measuring impressions).

Where the Audit Stalls

Now, a word of caution, because I’ve watched enough teams run this exercise to know exactly where it goes sideways.

The audit works. Teams map their claims against their corroboration, they find the gaps, and for about a week, everyone is energized. Then the earned gaps are handed to the PR team, the shared gaps to the social team, and everyone returns to the same silos that created the gaps in the first place. 

Six months later, the evidence base looks exactly like it does right now.

Because the audit tells you what’s missing. It doesn’t tell you who owns fixing it—and ownership is where this stalls, every single time. 

We saw it in a recent engagement with three strong differentiators on the website, and none of them confirmed anywhere that the client didn’t control.

The other thing teams underestimate is time. Earned credibility doesn’t move a model’s confidence score in a quarter—it compounds over years, the way trust always has. 

That’s not a reason to despair. It’s the reason to start before your competitors get another year of compounding you can’t buy back. 

Remember how I told you last week that the best time to plant a tree was 10 years ago, and the second best time is today? 

Same.

Plant your tree!

And if you’ve run the claims-versus-corroboration exercise and you’re staring at a list of unconfirmed claims, unsure which earned or shared play to run first—or who should own it—that’s a 30-minute call we’d love to have with you!

© 2026 Spin Sucks. All rights reserved. The PESO Model® is a registered trademark of Spin Sucks.