Because "AI visibility" is not one scoreboard. It is at least four separate ones, run by systems that disagree with each other about which sources to trust, how many of them to cite, and how closely a brand mention has to match an actual citation. Semrush's 2026 AI Visibility Index, built from 126 million real US prompts, found that only 36 brands worldwide held top-100 visibility on ChatGPT, Gemini, Google AI Mode and Google AI Overviews in every single month of the study. Everyone else, including plenty of large, well-known companies, was strong on one surface and close to invisible on another. If you are tracking a single "AI visibility score" and treating it as a verdict, you are almost certainly missing the half of the picture where you are losing.
What exactly did Semrush measure, and how big was the study?
Semrush's 2026 AI Visibility Index, published June 26, 2026, analyzed 126 million US AI search prompts collected between January and April 2026, and used that data to build benchmarks across 22 industries. The study covers four distinct AI surfaces: ChatGPT, Gemini, Google AI Mode and Google AI Overviews. That is a meaningfully larger and more current dataset than most of the citation studies that get quoted in this space, and Semrush tracked the same brands across all four surfaces over the same months rather than taking a single snapshot of any one of them.
That design choice is what makes the finding useful. A one-time ranking tells you who is cited most today. Tracking the same brands across four platforms for four straight months tells you whether being cited on one platform says anything about being cited on another. For most brands, it turns out, it does not.
How rare is it to actually be visible everywhere?
Rare enough that Semrush gave it a name. The report identifies a "Universal 36", the only brands that maintained top-100 visibility across ChatGPT, Gemini, Google AI Mode and Google AI Overviews in every month of the study. Out of however many brands were tracked across 22 industries, 36 managed consistent visibility on all four surfaces, all four months. The examples Semrush names are the kind of brand most people would guess anyway: YouTube, Google, Reddit, Amazon, Facebook, Apple, Walmart, Disney and Nintendo.
That list is instructive precisely because it is unsurprising. These are platforms and household names with an enormous, constantly refreshed public footprint: Wikipedia entries, news coverage, user-generated content, product reviews, and in several cases the fact that they own or power parts of the infrastructure the AI systems themselves run on. If universal visibility is reserved for brands at that scale, the realistic goal for almost everyone else is not "be visible everywhere" but "know exactly where I am visible and where I am not, and treat each one as its own campaign."
Source: Semrush's 2026 AI Visibility Index, published June 26, 2026, linked in the article.
Why does the same brand perform so differently platform to platform?
Because the platforms do not work the same way, and the gap starts with something as basic as how many sources each one bothers to cite. Semrush's data shows ChatGPT cites an average of 15 sources per response, while Gemini cites an average of just 3. A brand with a solid but not dominant public footprint has a real chance of making ChatGPT's wider net. The same brand has to be in Gemini's top handful of trusted sources for that topic, or it does not appear at all.
Gemini adds a second wrinkle that makes this worse than it sounds. According to the same study, the overlap between the brands Gemini mentions in its answer text and the domains it actually cites as sources can be as low as 30%. In other words, Gemini will often talk about a brand without linking to or crediting any of that brand's own content, which means a company can show up in the conversation while still showing zero footprint in the citation data a visibility tool reports back. Which websites ChatGPT, Gemini and Perplexity actually cite already showed that each platform pulls from a different shelf of sources; this data shows the gap runs even deeper than source selection, into whether being mentioned and being cited are the same thing at all.
Does this vary by industry, or is it the same everywhere?
It varies sharply, and the variation tells you something about how contested a category is. Semrush's concentration data shows the top three brands capture 82.9% of all mentions in News and Media, and 76.9% in Consumer Electronics, categories where a small number of globally dominant names already crowd out the field in ordinary search, let alone in an AI system trained to recognize prominence. Industrial sits at 42.2% and Finance at 41.4%, categories with a much longer tail of mid-sized, regionally relevant players that can realistically carve out visibility without unseating a handful of giants.
If you operate in a category that looks like Industrial or Finance in this data, the opportunity is real: AI systems are not yet locked into a short list of answers, and a well-documented, well-covered mid-sized brand can take a meaningful share of the mentions. If you operate in a category that looks like News and Media or Consumer Electronics, the realistic goal shifts from "win the general answer" to "own the specific, narrower questions" where the giants have not already saturated every citation.
Do most marketing teams even know which platform they are winning or losing on?
No, and that gap is the most actionable finding in the whole study. Semrush's accompanying marketer survey found 45% of marketing leaders say they cannot accurately measure their brand's visibility within AI-generated answers, and only 9% say they have tools that cover all the relevant metrics across platforms. That means the large majority of teams reading an "AI visibility" number today are reading a partial one, often from a single platform's data, without knowing it.
The same survey found a clear payoff for closing that gap: 81% of organizations that fully integrate SEO and AI-visibility work into one workflow reported increased traffic or leads from AI platforms, against just 36% among organizations that manage the two separately. AI visibility tools can tell you a brand is being ignored on a given platform; that 81% versus 36% split suggests the bigger gain comes from acting on what the tools show rather than from the tools themselves.
What should a brand actually do with this, platform by platform?
- Check each platform separately, not as an average. Ask ChatGPT, Gemini, Google AI Mode and a plain Google search the same question about your category, and write down which sources each one actually names. How to measure GEO properly means tracking per-platform results, not one blended score that hides exactly the kind of gap this study found.
- Expect Gemini to be the hardest surface and the least forgiving of a thin footprint. With only 3 sources cited per answer on average, Gemini rewards being one of the handful of most-trusted domains on a topic, which usually means sustained earned coverage and a well-maintained Wikipedia presence, not a single strong article.
- Treat mentions and citations as two different metrics on Gemini specifically. A brand name appearing in the answer text is not the same as being linked as a source, and a visibility tool that only counts one of those will overstate or understate your real position.
- Weight your effort by category concentration. Getting cited in Google's Gemini answer for a narrow, specific industry question is a realistic goal in a fragmented category like Finance or Industrial; in a concentrated category like Consumer Electronics or News and Media, aim at the specific sub-questions the dominant brands have not already answered.
- Build the cross-functional process the 81% figure points to. Pair whoever owns SEO with whoever owns PR and AI-visibility tracking, so a new piece of earned coverage gets checked for its effect across all four surfaces, not just folded into a general SEO report.
The direct answer, for anyone who read this far
A brand can genuinely dominate ChatGPT's answers and barely register on Gemini, because the two systems cite a different number of sources from a different pool, and Gemini in particular often mentions a brand without citing it at all. Semrush's data shows only 36 brands worldwide, almost all household names, achieved consistent visibility across ChatGPT, Gemini, Google AI Mode and Google AI Overviews over four straight months. For everyone else, the real work is tracking each platform on its own terms and closing the specific gaps that show up, rather than chasing one combined score. If you want help building that platform-by-platform picture, get in touch with our team.



