Digital PR for SEO and AEO: Earning the Mentions AI Trusts

Digital PR for SEO means earning coverage and mentions from journalists, not building links directly. It still helps rankings, but its bigger job now is answer engines: Ahrefs found brand mentions correlate with AI visibility three times more strongly than backlinks. AI models decide who to trust from mentions like these, and digital PR is how you earn them.

Quick links

• Why does digital PR matter more now than it used to?
• Why do AI answer engines trust mentions over backlinks?
• What counts as a mention that AI actually notices?
• What does a mention-earning digital PR campaign look like?
• How do you know if it’s working?
• What’s a realistic timeline?
• FAQ

Why does digital PR matter more now than it used to?

Digital PR used to be a link-building tactic with a nicer name. Now it’s closer to infrastructure: the thing that decides whether AI knows you exist at all.

For a decade, the pitch for digital PR was straightforward: get a journalist to cover a data story, pick up an editorial backlink, watch domain authority climb. That still works. In a Q1 2026 survey of 500 SEO professionals, Reporter Outreach found digital PR was rated the best-performing link-building tactic by 34% of respondents, nearly double the next method, guest posting, at 18%.

What’s changed is what those mentions do beyond the ranking boost. AI models triangulate credibility from what other sites say about you, not just what your own site says. A feature in an industry publication, a quote in a trade press piece, a listing in a respected directory. Every one of those is now doing double duty, feeding both the search engine and the answer engine at once. Our guide to what AEO actually is covers why that off-site signal matters as much as anything on your own pages.

Why do AI answer engines trust mentions over backlinks?

Because a mention tells an AI model what to trust, while a link only tells it where to go. Large language models extract entities and context from text; they don’t need to click anything to register that a source is talking about your brand.

The clearest evidence for this comes from Ahrefs, which analysed 75,000 brands to find which signals best predicted appearing in Google’s AI Overviews. Branded web mentions correlated at 0.664. Backlinks scored 0.218, roughly a third as strong. Branded search volume and branded anchor text also outperformed raw link counts. The study’s own conclusion is blunt: the number of pages you publish barely moves the needle, but being talked about does.

That pattern holds beyond Google. Muck Rack’s ongoing “What Is AI Reading?” research has now tracked three editions since mid-2025, and the finding hasn’t moved. In the May 2026 edition, earned media accounted for 84% of the more than 25 million citations Muck Rack analysed across ChatGPT, Claude and Gemini responses. Paid and advertorial content made up 0.3%. Journalism alone was 27% of everything cited. AI models are, in effect, reading the news and trusting it more than they trust anything a brand publishes about itself.

That’s the whole case for digital PR in one sentence: you cannot buy your way into an answer, but you can earn your way in.

What counts as a mention that AI actually notices?

Almost any clear, in-context reference to your brand name, whether or not it links back to you. The bar is lower than most people assume, but the context still has to be specific.

A few forms that count:

• A named mention in an article, even without a link. “Aeonix, a Lancashire-based agency, found that…” registers as a data point.
• A quote attributed to a named person at your company, which ties your brand to expertise on a specific subject.
• A listing in a directory or “best of” roundup, which is exactly the kind of comparison content AI models lean on for recommendation-style queries.
• A citation in someone else’s research or report, which carries unusual weight because it’s a third party vouching for your data.

What doesn’t count for much: a mention buried in a sponsored post, a link farm placement, or anything that reads as paid. The Muck Rack figures above make that distinction sharply. AI systems are unusually good at telling earned coverage apart from advertorial, and they weight accordingly.

What does a mention-earning digital PR campaign look like?

In practice, three repeatable moves: give journalists something to report, make your people available as sources, and keep your listings clean. None of them are exotic.

Original data and survey-led stories

Journalists need something new to write about, and a genuine data point from your own work or a small original survey is the easiest way to give it to them. This doesn’t need to be a huge study. A pattern from your client work, a survey of 100 people in your sector, a before-and-after from a project you shipped, all of these are usable if the number is real and the framing is honest. The dataset from Ahrefs cited earlier is itself a textbook example: a company ran an analysis nobody else had, and every SEO publication on the internet ended up covering and linking to it.

Expert commentary on stories already running

Reporters covering a live story often need a quotable expert fast. Positioning your named specialists to respond to relevant news, through direct pitching or a journalist-request platform, earns a mention tied to real-time context, which is exactly the kind of fresh, specific reference AI systems favour over generic evergreen copy. This is slower to compound than a big data story but far more consistent, since there’s always a story running in most sectors.

Listings, directories and comparison content

The unglamorous end of digital PR, and one of the highest-leverage. Being included, accurately and consistently, in the directories, “best of” lists and comparison sites your buyers and AI models both check is a mention that keeps earning long after the outreach email is sent. It also reinforces the entity consistency that both classic SEO and AEO depend on: the same name, description and category, repeated everywhere your business is listed.

How do you know if it’s working?

Track whether AI engines mention you before and after a campaign, not just whether a journalist ran your story. That’s the outcome digital PR is now being asked to move.

The simplest method is manual: ask ChatGPT, Perplexity and Gemini the buyer-style questions your prospects would ask, before a campaign and again a few weeks after coverage lands, and note whether you start appearing or whether the description of your business gets more accurate. Our piece on LLM SEO goes into how models select sources in more depth, and our AI search optimisation field guide covers monitoring across the full set of surfaces, AI Overviews, AI Mode, ChatGPT and Perplexity, rather than one at a time.

Keep an eye on the mundane metrics too: branded search volume, direct traffic and referral traffic from the publications you’re covered in. All three showed up as meaningful correlating factors in the Ahrefs study, alongside raw mentions, so a lift in any of them is a reasonable proxy that the coverage is landing with real people as well as with models.

What’s a realistic timeline?

Faster than classic link building for the first signal, slower for the pattern that actually moves recommendations. Expect weeks for the former and months for the latter.

Some AI systems, particularly Perplexity and search-grounded modes of ChatGPT, retrieve fresh web content close to real time, so a single strong placement can show up in an answer within days or weeks of publication. Training-data-based answers and the deeper entity picture Google and other models build up move slower. That needs a pattern: several credible mentions over a few months, not one story, before a model treats your brand as an established answer rather than a one-off reference. Treat digital PR as an ongoing programme, the same way you would treat content or technical SEO, rather than a single campaign with a defined end date.