Do brand mentions impact visibility in AI search? What the 2026 data shows
Brand mentions correlate with AI search visibility about three times more strongly than backlinks. Here's what the data says, where it's thin, and what to do.
Written byMuhammad Usman SiddiquiFounder and editor-in-chiefWhy you can trust The AI Future
Our writers research every product from its official pages, independent tests and user reports, link each claim to its source and correct mistakes openly. Unless an article says otherwise, we haven't tested the product hands-on ourselves. Companies can't pay for coverage or see reviews before they're published.
Read our editorial policy
Yes, brand mentions do seem to impact visibility in AI search, and they look like one of the strongest signals we can measure. In an Ahrefs study of 75,000 brands, how often a brand was mentioned across the web had a 0.664 correlation with how often it showed up in Google's AI Overviews, roughly three times the figure for backlinks (0.218). That's correlation, not proof of cause, but it's a big enough gap that I'd take it seriously.
So the honest answer is "probably, and more than links". The rest of this piece covers why that might be, where the evidence is thin, and what you can actually do about it.
The short answer
AI search tools like ChatGPT, Gemini and Google's AI Overviews tend to name brands that the wider web already talks about. If people write about you on review sites, forums, news outlets and Wikipedia, you're far more likely to be mentioned in an AI answer. If they don't, your own site's content probably won't be enough on its own.
The mechanics behind AI brand visibility
Here's a picture that helps me. A new cafe opens in a small town. The owner can put up a lovely sign (that's your website), but when a visitor asks a local "where should I get coffee?", the answer depends on what the local has heard: what neighbours said, what the paper wrote, what people grumbled about online. An AI model is a bit like that local. It doesn't visit every shop; it repeats the town's chatter.
There are two layers to that chatter. The first is training data. Large language models learn from huge amounts of web text, so a brand that appears often, in the right context, becomes part of what the model "knows". Ryan Law at Ahrefs, quoted in the same study, says unlinked mentions "have very little impact on SEO, but a much bigger impact on GEO." (GEO is generative engine optimisation, a term from a 2023 Princeton-led paper that found certain content changes could lift visibility in generative answers by up to 40%.)
The second layer is live retrieval. When you ask a question, many AI tools run fresh searches and pull in pages to ground the answer. Google says AI Overviews and AI Mode may use "query fan-out", issuing several related searches across subtopics before writing a response. Each of those searches is another chance for a third-party page that mentions you to get picked up.
Put the two together and the logic's fairly simple. More mentions in more places means more chances to be learned and more chances to be retrieved.
Which ai brand visibility factors matter most, then? The Ahrefs numbers are the clearest public dataset I've found, so here they are in one place (Spearman correlations with AI Overview brand mentions):
| Factor | Correlation | Type |
|---|---|---|
| Branded web mentions | 0.664 | Off-site |
| Branded anchors (brand name in link text) | 0.527 | Off-site |
| Branded search volume | 0.392 | Off-site |
| Domain Rating | 0.326 | Links |
| Referring domains | 0.295 | Links |
| Branded organic traffic | 0.274 | Traffic |
| Backlinks | 0.218 | Links |
| Branded ad traffic | 0.216 | Paid |
The top three are all about how people talk about you elsewhere. Paid search barely registers.
Real examples from the citation data
The most useful way to answer "which websites influence what AI says about my brand?" is to look at which sites AI tools actually cite. Several trackers publish this now, and they mostly agree.
In Similarweb's analysis of nearly 600,000 citation events from January to February 2026, Wikipedia took 13.15% of ChatGPT's US citations and Reddit took 11.97%. OpenAI's own site came third at 6.21%, which surprised me a little. Google AI Mode looked different: Fandom led with 7.16%, followed by Wikipedia, YouTube and Reddit, and google.com itself sat fifth.
Peec's numbers point the same way. Its study of 30 million cited sources across ChatGPT, AI Mode, Gemini, Perplexity and AI Overviews (March 2026) put Reddit first, then YouTube, LinkedIn and Wikipedia, with G2, Yelp and TechRadar also in the top ten. And Ahrefs' monthly tracker, which it updates automatically, has Wikipedia at 10.1% of ChatGPT citations in November 2026.
Industry matters a lot, though. Peec notes that in hotels, Booking.com and Airbnb often outrank Reddit. For a B2B software company, I'd guess G2 and LinkedIn carry more weight than Fandom ever will.
The Ahrefs quartile data is the starkest example. Brands in the top 25% for web mentions averaged 169 AI Overview mentions; the next quartile averaged just 14, and the bottom half averaged 0 to 3. About 26% of the brands studied had no AI Overview mentions at all.
Benefits and limitations
The appeal is obvious. Mentions are something you can influence without owning the page: a podcast interview, a review on G2, a Reddit thread where a real customer recommends you, a news story about your funding round. None of it needs a link to count, as far as we can tell from the data. That's good news for smaller brands that have struggled to win backlinks.
Now the caveats, and they're real. Correlation isn't causation, and Ahrefs says so plainly. Big brands get mentioned a lot because they're big, and they'd probably show up in AI answers for the same reason; mentions might be a symptom of popularity rather than its cause. The Ahrefs study also only looked at sites with a Domain Rating above 40, so we know less about tiny brands. And the citation studies measure different things over different months, which is why their rankings don't match exactly.
There's a fragility issue too. Peec warns that both Google and ChatGPT have pulled back from Reddit in the past, so a brand whose whole presence lives in one community carries concentration risk. I think that's the most underrated point in all of this research.
Finally, mentions aren't always flattering. If the most-discussed thread about your product is a complaint, AI tools can repeat that too.
Practical takeaways for your brand
Start by finding out where you stand. Ask ChatGPT, Gemini and Perplexity the questions your customers ask, and note who gets named. Tools like Ahrefs Brand Radar, Peec and Similarweb do this at scale; we've rounded some up in our guide to AI brand visibility analysis tools.
Then go where the citations are. For most categories that means Wikipedia (if you genuinely qualify), Reddit, YouTube, LinkedIn, and the review sites in your niche. Don't astroturf Reddit, though; Peec's advice is to participate only authentically and add real value.
Earn coverage the old-fashioned way. Original data, expert commentary for journalists, and partnerships tend to get your name written down by people who aren't you. Honestly, this is just good PR with a new reason to do it.
Don't drop SEO. Google is clear that no special optimisation is needed for AI features: a page has to be indexed and eligible for a snippet, and normal best practice applies. Since AI Overviews pull from search results, ranking still helps you get cited. Our explainers on what AI Overviews are and our AI Mode review cover how those features pick sources.
Keep your facts consistent. If your pricing, founding date or product names differ across your site, LinkedIn and your Google Business Profile, you're giving models mixed signals.
Next up: ads, agents and shifting sources
Paid placements are arriving, which could change the maths. OpenAI said in January 2026 that it would test ads at the bottom of ChatGPT answers for free and Go users in the US, and it promised that "Ads do not influence the answers ChatGPT gives you." By April, LLMrefs found ads in 19.6% of sampled US ChatGPT responses, with HubSpot the biggest buyer it detected. For now the ads sit beside the answer, not inside it, so organic mentions still decide who gets named.
I'd expect two other things. Citation sources will keep moving month to month (Ahrefs' own table shows several domains entering its top 50 for the first time in November), and agent-style AI that books or buys things will make being named even more valuable. Will mentions stay the strongest signal? Probably, as long as models learn from text. But I wouldn't bet a whole marketing budget on any single ranking factor that's only a couple of years old.
My view, for what it's worth: treat brand mentions as your main lever for AI visibility, measure them monthly, and keep the SEO basics going.
FAQ
Which websites influence what AI says about my brand?
Mostly the sites AI tools cite most: Wikipedia, Reddit, YouTube, LinkedIn and review platforms like G2 and Yelp, according to Similarweb and Peec data from 2026. Your industry shifts the list, so check which domains get cited for your own customer questions before deciding where to focus your effort.
Who helps brands build off-site signals for AI visibility?
Usually digital PR firms, SEO agencies with a GEO or AI search offer, and community or review-management specialists. Some in-house teams do it with monitoring tools like Ahrefs Brand Radar or Peec. If you're comparing providers, our list of AI SEO agencies is a reasonable starting point; ask any agency for before-and-after mention data.
What are the best ways to analyze brand traffic from AI search engines?
In Google Analytics 4, check the built-in AI Assistant channel, as Loves Data explains, or build a custom channel group with a regex for sources like chatgpt.com and perplexity, which Ahrefs walks through. Visits from AI Overviews and AI Mode are counted as organic search in Search Console, so you can't separate them cleanly. Ahrefs also estimates LLMs send only about 0.1% of traffic, though it thinks that's an undercount because some AI platforms withhold referrer data.
Do unlinked brand mentions count for AI search?
The evidence suggests they do. Ahrefs counted linked and unlinked mentions together and got its strongest correlation (0.664). Models learn from text, so a plain mention likely still helps. It isn't proven causation, though.
Do backlinks still matter for AI visibility?
Yes, but less than you might expect. Backlinks correlated at only 0.218 with AI Overview mentions in the Ahrefs study, and Domain Rating at 0.326. They still help you rank in Google, and ranking pages feed AI Overviews, so I wouldn't stop link building. I'd just stop treating it as the main lever.
