Why Earned Media Matters in ChatGPT, Claude, Gemini, and Perplexity

Profound’s Nick Lafferty explains what AI search is actually citing, why earned media matters more than most marketers realize, and why PR teams need to start looking at citations — not just coverage

Every few months, marketing invents a new acronym that makes normal people want to close their laptop. And the acronym of 2025 and continuing into 2026 is AEO — Answer Engine Optimization.

You may also see it called GEO, AI visibility, AI search optimization, LLM visibility, or seventeen other things that while there are nuanced differentiators, they all basically mean the same thing… when someone asks ChatGPT, Claude, Gemini, Perplexity, AI overviews or another AI tool a question about a brand, does your brand show up?

And perhaps MOST importantly for us publicists — what information is AI citing for these inclusions? AKA is our work making a difference in the new era of search?

Because for years, PR teams have been told they are “top of funnel.” Important, yes. Valuable, yes. But often harder to measure than paid search, affiliate, or performance marketing. Don’t get me started on UVM.

AI search changes some of that. Not because it magically solves attribution. It does not. Instead, it creates a new layer of visibility where third-party credibility matters a lot. If AI tools are pulling from outside sources to answer questions about products, brands, categories, comparisons, and recommendations, then earned media is not just a nice awareness win. It can become part of the source material shaping the answer.

To understand this better, I spoke to Nick Lafferty, founding marketing engineer at Profound, a platform that helps companies understand, control, and improve how they show up in AI search.

Quick side note, Nick comes from affiliate and publishing himself. So he understands the weird little overlap between content, performance, visibility, and revenue that many of us live in every day.

Here is the version of AI search visibility that publicists actually need to understand.

First, what is Profound?

Nick’s simplest explanation is that Profound helps companies understand, control, and scale how they show up in AI search.

Profound tracks what AI tools say when people ask questions in a category.

It looks at things like:

• Is the brand mentioned?

• Is a competitor mentioned instead?

• Which brand appears first, second, or third?

• Which sources did the AI cite or rely on?

• Is the brand’s own website being cited?

• Are third-party articles, Reddit threads, blogs, YouTube videos, or other sources influencing the answer?

• What changed over time?

That is the “visibility” side.

The other side of Profound is action. The platform also helps marketers build AI agents that can monitor, research, draft, update, and automate repetitive workflows tied to AI visibility.

For a PR person, the most important aspect of a response isn’t what is said, but what and WHO is being cited.

Because if you know which sources AI tools are using to answer questions about your client’s category, you can start to understand which press hits actually matter in the AI layer of our work. Sure, you will still have coverage that might LOOK impressive in an EOM report, but they had little to no impact on how the brand is represented in AI answers.

Why this matters for PR specifically

According to Nick, across the AI search citations Profound has tracked, only about 5% of citations come from a brand’s own domain.

That means roughly 95% come from somewhere else.

Third-party coverage. Publisher articles. Blogs. Social platforms. Reddit. LinkedIn. YouTube. Review sites. Comparison pages. Industry lists. Earned media.

That does not mean every press hit matters equally. But it does mean PR is not a side conversation in AI visibility. It’s central to it.

For years, PR has been trying to prove that earned media influences how people discover, understand, compare, and trust brands. AI search is basically turning that into a new kind of data problem.

If ChatGPT or Perplexity is answering “best modular sofas for families,” “best skincare brands for sensitive skin,” “top affiliate platforms for agencies,” or “best PR agencies for affiliate marketing,” the question becomes what sources does the model use to form that answer?

And did your PR work help create or influence those sources?

What “AEO” actually means

AEO stands for Answer Engine Optimization.

This is the work you do to make sure your brand is recommended, mentioned, cited, or accurately represented when someone asks an AI tool a question.

That is it.

It is not just SEO with a new coat of paint. And it’s also not just PR either.

It sits in the messy middle of:

• PR

• SEO

• content

• affiliate

• product marketing

• reputation management

• competitive intelligence

• customer reviews

• community conversations

Which is why publicists should care, but also why publicists should not pretend they own the whole thing alone.

This is going to require PR teams, SEO teams, content teams, affiliate teams, and brand teams to work much more closely together than they have in the past.

Annoying? Yes. True? Also, yes.

Where AI answers are getting information

A common misconception is that AI tools are mostly pulling from a brand’s website. And sometimes they are. But according to Profound’s data, most citations are coming from outside the brand’s owned domain.

That includes:

• editorial articles

• publisher reviews

• blogs

• forums

• Reddit threads

• LinkedIn posts

• YouTube videos

• comparison content

• category pages

• third-party lists

• review sites

• other public web content

For publicists, this is a huge shift in how we think about coverage. Traditionally, a press hit had a short life cycle:

You pitch it. > It runs. > You share it. > Maybe the client posts it on social. > Maybe sales uses it. > Maybe it gets added to an “as seen in” section.

Then everyone moves on.

But in AI search, a good third-party source may keep influencing answers long after the initial press cycle.

The inverse is also true. A press hit you are very proud of may not be influencing AI answers at all. (ehhhmmmm That New York Times story.)

That is the uncomfortable part and this is the messy middle of where we are as publicists and marketers in 2026 with the search landscape evolving.

Does a tier-one placement move the needle?

The honest answer is sometimes.

Nick was very direct about this - A tier-one placement does not automatically improve AI visibility.

Why? Because some major publications block AI crawlers or limit how AI platforms can access their content. The New York Times, for example, has been one of the most visible examples of a publisher restricting AI crawler access.

So yes, a major media hit may matter enormously for credibility, investor relations, client happiness, sales enablement, and plain old human awareness.

But that does not mean it is automatically being cited by ChatGPT, Claude, Gemini, or Perplexity.

Nick shared a visual showing that, in Profound’s dataset, the vast majority of earned media citations come from non-tier-one sources, not the big marquee publications PR teams tend to celebrate most.

The takeaway is not “tier-one press does not matter.” It’s just that top tier press does not automatically mean AI is using those as sources.

A publicist may still want The New York Times, Vogue, Forbes, Good Morning America, or The Wall Street Journal for very good reasons.

But if the specific goal is to influence AI answers, you need to know whether that outlet, page, format, or syndicated version is actually accessible and cited by the models.

That is a different question than “is this a prestigious outlet?”

What types of coverage may punch above their weight?

This is where it gets less clean, because the answer depends on the model and the category.

Nick gave a helpful example: Google’s AI products tend to prefer Google-owned or Google-adjacent data. Gemini, AI Mode, and AI Overviews are more likely to cite or surface YouTube than ChatGPT is.

So if you are optimizing for Gemini or Google AI Overviews, YouTube may matter more than you would assume.

But if you are looking at ChatGPT, that may not be the same playbook.

The broader point is that there is no universal list of “best outlets for AEO.”

It depends on a variety of factors:

• the client’s category

• the prompt being asked

• the AI model being used

• the sources that model tends to cite

• the competitors already winning those answers

• whether the source is accessible to AI crawlers

• whether the content is specific, useful, and structured enough to answer the question

That is a little frustrating, but it is also where the opportunity is.

The teams willing to actually look at the data will have an advantage over the teams still trying to apply a generic media list to a new discovery channel.

What coverage gets ignored?

Nick called out one format publicists should pay attention to which is the traditional wire distribution.

PR Newswire, GlobeNewswire, and similar wire services are not necessarily being cited often in AI search. According to Muck Racks Generative AI report from December 2025, Press Release citations are up 5x but the devil is in the details and to quote one of my fav publicists Carly Martinetti, “its just a whopping 0.2% to 1% of total AI citations for newswire services, and from 1.2% to 6% for all other press release citations. Journalism still drives 20-30%.” (Follow her on LinkedIn)

That does not mean press releases are useless. They still serve a function, and I hopefully have a Q&A coming on the topic very soon. They have case studies for all areas of business including feed investor communications, support search, provide clean approved messaging points, and more. And yes, they are rising in their citations.

I say all that to say, if your AI visibility strategy is “we put out a press release,” that’s not enough. AI tools appear to prefer useful, specific, third-party content that helps answer the question being asked.

Want to try Profound? What should PR teams look at first?

Citations - A brand mention tells you whether the AI included your client in the answer. A citation tells you what source helped shape that answer.

Nick recommends looking at citations over a period of time. For example, seven days or fourteen days, which allows you to see which pages and domains are consistently influencing AI answers.

From there, PR teams can start asking much better questions:

• Which outlets or domains are showing up again and again?

• Are our earned media hits being cited?

• Are competitors being cited from sources where we are absent?

• Are there category guides, reviews, or comparisons we should be pitching?

• Are there specific pages that are carrying a competitor’s visibility?

• Are there content gaps our client needs to fill?

• Are AI answers citing outdated information?

• Are the answers accurate?

• Are they favorable?

• Are they missing key proof points?

This is where AI visibility becomes useful for PR. Not as a vanity metric but a stratgic roadmap.

How to use competitor visibility

One of the most useful parts of Profound is that it shows not just how your brand appears, but how competitors appear.

Nick said one of the most common “aha” moments is when a brand realizes a competitor has one specific page, article, or resource that is doing a lot of work in AI search.

The client may have had no idea that outlet/page mattered. But once they see it being cited, it becomes another dot on your roadmap.

For PR and content teams, that can turn into questions like:

• Do we need a better owned resource on this topic?

• Do we need more third-party validation around this product feature/claim?

• Do we need review coverage in a category where competitors are already showing up?

• Do we need comparison content?

• Do we need to update an existing page that used to perform but has started dropping?

• Do we need to pitch the sources AI is already relying on?

This is also very relevant to affiliate and commerce PR.

Because a lot of high-intent AI prompts sound very similar to commerce queries:

• best sofa for families

• best luggage for international travel

• best hair dryer for curly hair

• best affiliate network for small brands

• best CRM for agencies

• best mattress for side sleepers

Those are not fluffy awareness questions. Those are buying-journey questions.

And if AI tools are answering them with citations from publishers, review sites, comparison pages, and affiliate-driven commerce content, PR teams need to understand which sources are shaping those recommendations.

Okay, so where does PR end and SEO begin?

This is one of the biggest questions for agencies.

If AI visibility depends on earned media, owned content, technical accessibility, structured information, reviews, Reddit, YouTube, and competitive content, who owns it?

PR?

SEO?

Content?

Affiliate?

The answer is most likely that no one team can own it alone.

Nick described AEO as a place where the old marketing silos start to blur. PR, SEO, content, product marketing, and affiliate all overlap because AI answers are not pulling from one neat channel.

Profound’s perspective is that companies may need a new kind of role to manage this work: the marketing engineer.

This does not mean every PR agency needs to hire a software engineer tomorrow. But it does mean your work needs to become more systems-driven and far less spray and pray. Using the same old media list for every single pitch.

The publicist of the future may not just be pitching stories. They may also be looking at citation data, identifying source gaps, monitoring competitor visibility, collaborating with SEO on owned pages, and using agents to track when a client’s AI visibility changes.

That is not traditional PR. But neither was affiliate PR ten years ago.

And yet here we are.

The manual version if you are not ready for a platform

Nick was blunt here - the manual version is painful.

But it does exists.

If you are not ready to use a platform, you can start by manually tracking important prompts in a spreadsheet.

For example:

• “What are the best [category] brands?”

• “What is the best [product] for [use case]?”

• “Which [brand/category] is best for [audience]?”

• “What are the top alternatives to [competitor]?”

• “Is [brand] good?”

• “What are the pros and cons of [brand]?”

• “Which [category] brands are recommended by experts?”

Then test those prompts across ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, or whichever platforms matter for your client.

Track:

• whether your client is mentioned

• which competitors are mentioned

• the order of mentions

• the sources cited

• whether the answer is accurate

• whether affiliate publishers, review sites, Reddit, YouTube, or media articles are influencing the answer

• what changed week over week

This is not perfect. It is also not scalable.

But it will teach you how AI answers are representing your client’s category.

And that alone will make you more useful than the person saying “AI search is important” without being able to explain what to do about it.

Where this is going

Profound is clearly not just thinking about dashboards.

Nick said the future is action.

The next phase is not just showing marketers that visibility changed. It is helping them do something about it.

Imagine a system that notices a competitor updated a page, finds your client’s equivalent page, drafts suggested updates, stages them in the CMS, and pings the PR and content teams in Slack for approval.

Or an agent that monitors whether key earned media citations dropped week over week.

Or one that alerts a publicist when a competitor starts showing up more often for a high-value AI prompt.

That is the direction this seems to be heading: not just AI visibility reporting, but AI visibility workflows.

For publicists, that matters because the job is going to get more technical whether we like it or not.

The good news is that PR people already understand one of the hardest parts: influence.

We understand which third-party sources carry credibility. We understand how narratives move. We understand why one sentence in the right article can matter more than ten owned blog posts.

The next step is learning how that influence shows up inside AI answers.

My take for publicists

You do not need to become an SEO overnight. You do not need to pretend every client needs an prolific “AEO strategy” because the acronym is hot.

But you do need to understand that AI tools are becoming a discovery layer that is increasing. Exponentially.

People are using them to research products, compare brands, ask for recommendations, summarize categories, and make decisions before they ever land on a website.

That means PR teams should start asking a new set of questions:

• When someone asks an AI tool about my client’s category, does my client show up?

  • If not, who does?

• What sources are shaping that answer?

  • Are those sources earned, owned, affiliate, social, Reddit, YouTube, or something else?

• Are our best proof points showing up?

• Are competitors getting credit for narratives we should own?

• Are our press hits actually being cited?

• Are we pitching the kinds of stories that AI tools can use?

This is not about replacing traditional PR. It is about adding another layer to the job. (I know, I know. Another layer?)

But look at it this way, it gives us MORE ways to prove our value, worth and a stratgic place for our work to move the needle for the brand. After all, visitors arriving via AI search platform citations (like ChatGPT, Perplexity, or Google AI Overviews) convert at 3-5x (and even higher for b2b) the rate of standard Google organic search.

---

Resources

Profound has a free Profound Index where brands can see how they compare in AI search visibility across their category. It includes competitor context, co-mentions, topic clusters, and mention position analysis.

Nick Lafferty
Email: nick@tryprofound.com
Founding Marketing Engineer @ Profound