Web mentions track AI Overview visibility three times as closely as backlinks. What CPG brands should chase

Ask a CPG founder what drives search visibility and you will often hear about links: the press hit with a link back, the blogger roundup, the guest post. For blue-link search that was a reasonable model. For Google's AI Overviews, the best data we have says links matter much less than being talked about.

The data comes from Ahrefs, which studied 75,000 brands to see which signals line up with how often AI Overviews mention a brand. Search Engine Land revisited the findings on November 6 in a piece on why authority alone no longer protects big brands. We think the numbers deserve a slower read from a packaged goods point of view, because the kind of mention a CPG brand earns is different from the kind a software company earns.

How the study was built

Ahrefs took domains with a Domain Rating above 40 and, for each, its highest-volume keyword with at least 800 monthly searches, which it used to identify the brand. It then searched millions of AI Overview responses for mentions of those brands using its Brand Radar tool. About 26% of the brands had zero mentions; the correlations were run on the remaining 74% or so. It used Spearman rank correlation, where 1 means two measures rise in perfect step and 0 means no relationship.

Two caveats before the numbers. First, this is correlation, and Ahrefs says so: brands that are widely discussed also tend to have more links, more searches and more content, so no single factor can be credited. Second, the study is general. It is not a CPG sample, and Ahrefs does not break it out by category.

The correlations

Show the data

Factor	Correlation
Branded web mentions	0.664
Branded anchor text	0.527
Branded search volume	0.392
Domain Rating	0.326
Referring domains	0.295
Branded organic traffic	0.274
Number of backlinks	0.218
Branded ad traffic	0.216
Branded ad cost	0.215

Source: Ahrefs, AI Overview brand visibility study. Correlation, not cause. Values from 0 to 1; higher means the factor rises more consistently with AI Overview mentions.

Three things stand out.

Mentions beat links by about three to one. Branded web mentions, meaning the brand's name appearing anywhere on the web, linked or not, correlated at 0.664. The number of backlinks correlated at 0.218. Divide one by the other and mentions track AI Overview visibility about three times as closely, which is our calculation and also the comparison Search Engine Land draws.

Links still count when they carry your name. Branded anchor text, where the clickable words of a link are the brand name, came second at 0.527. So links are not irrelevant. A link that says your brand's name is a mention as well as a link. A link that says "click here" or "this product" is much less of either.

Paid search does not buy AI Overview presence. Branded ad traffic and branded ad cost sat at the bottom, at 0.216 and 0.215. Spending on brand keywords and appearing in the AI summary for your category appear to be different things.

The visibility cliff

The second finding is starker. Ahrefs split the brands into quarters by how many web mentions they had and averaged their AI Overview mentions.

Show the data

Web-mention quartile	Average AI Overview mentions
Top 25%	169
Second 25%	14

Source: Ahrefs, AI Overview brand visibility study. Ahrefs reports the bottom two quartiles averaged between 0 and 3 mentions; not charted because no exact values were given.

The top quarter averaged 169 AI Overview mentions. The next quarter averaged 14. The bottom half averaged somewhere between zero and three. That is not a slope. It is a cliff, and most brands are below it. Ahrefs' own summary is that if your brand sits in the lower half of web mentions, it is essentially invisible to AI Overviews.

Search Engine Land added a related number from Ahrefs' separate citation research: in July 2025, 76% of URLs cited in AI Overviews also ranked in Google's organic top 10; a March update with better citation detection put it at 38%. Ahrefs says part of that change comes from the better detection, so the two are not directly comparable. The direction still matters. Ranking well is no longer close to the whole story.

Factor in the study	Correlation	The CPG version of it
Branded web mentions	0.664	Your name on retailer listings, review sites, recipe and how-to pages, Reddit threads, trade press, local news
Branded anchor text	0.527	Links to you that use the brand name, such as a retailer's "Shop Brand" link or a press link on your name
Branded search volume	0.392	People typing your brand name into Google, often after seeing it on shelf or in social
Domain Rating	0.326	The overall link strength of your own domain
Number of backlinks	0.218	Raw link count, regardless of what the link says
Branded ad cost	0.215	Money spent on brand keywords

Source for the correlations: Ahrefs. The right-hand column is our reading.

Why this fits what else we know

This lines up with two studies we covered last week. Babel42's research found that only 15% of AI brand recommendations cited the brand's own website; the rest came from other people's pages. And our think piece on the brand website as the receipt, not the reason argued that a brand's own site confirms facts while third parties make the case. Ahrefs' correlations are the same picture from a different angle: the web talking about you is what tends to go with being named.

It also fits how Google describes AI Overviews. Google's documentation for site owners says both AI Overviews and AI Mode may use a "query fan-out" technique, issuing multiple related searches across subtopics to build an answer. A brand that is discussed in many places has more chances to turn up in one of those searches.

What a CPG brand should actually chase

The tempting response is to tell the PR agency to stop chasing links and start chasing unlinked mentions. Search Engine Land warns against exactly that swap, and we agree. Mentions without anything citable on your own site leave an answer engine knowing you exist with nothing specific to say. Here is what we would do instead.

Aim for mentions next to the problem, not just the name. A mention that says your brand's name in the same sentence as "fragrance-free," "for toddlers," or "low sugar" ties the brand to the question a shopper asks. Search Engine Land calls this topical proximity. A bare name in a list of "brands we love" ties you to nothing.

Treat retailer pages as the biggest mention engine you have. For most packaged goods brands, the pages that name the brand most often are retailer listings and their reviews. A complete, accurate listing on every retailer you sell through, as tip 30 describes, is a mention program in itself.

Feed the places people discuss the category. Recipes, how-to guides, parenting forums, Reddit. Tip 9 covers how to show up on Reddit without spamming it. Sending product to reviewers and testers who publish honest results does more than a sponsored roundup.

Make the press link carry your name. When you do earn coverage, ask for the link on the brand name rather than a generic phrase. Branded anchors were the second-strongest factor.

Publish something worth citing. Original data, a real survey, a clear explainer on an ingredient question your category gets asked. When others cite it, you earn the mention, the link and the topic association at once. Tip 11 covers getting into the sources AI answers cite.

Keep the name consistent. A brand that appears as three different spellings across retailers and press is three weaker entities. Tip 18 on helping AI see your brand as one entity is the groundwork.

How to check where you stand

You do not need Ahrefs' dataset to get a rough read. Search the web for your brand name in quotes and note where it appears outside your own site and retailers. Do the same for your two closest competitors. Then ask Google the five category questions you most want to be in the answer for, and note which brands the AI Overview names. If the brands that appear are the ones with the most outside discussion, you have the same pattern Ahrefs found, in your own aisle.

For a repeatable version, some teams track mentions in an SEO tool and track answers separately with something like SearchPanel, the AI visibility product from XENA Intelligence, which asks assistants shopper-style questions and records how products appear. The useful comparison is the two lines side by side over months.

The limits

This is one study, of correlation, across all kinds of brands, using one vendor's mention index and its own definitions. The brand set was filtered to domains with a Domain Rating above 40, so very small brands are not represented at all. And AI Overviews change; the study predates some of this year's changes to where they appear and how often.

What we would take from it is a priority order. If your team has one month and one budget line for AI visibility, the evidence points toward being discussed, by name and in context, in more of the places shoppers and reviewers already write. Links are part of that. They are no longer the point of it.