We Analyzed 50 Brands in ChatGPT, Claude and Perplexity — Here's What We Found
We expected the results to roughly mirror Google. They didn't.
We ran 50 brands across four AI assistants, asking the buyer questions their customers actually type, and the pattern that emerged should worry any agency still reporting on rankings alone. Domain authority barely predicted who got recommended. Traffic barely predicted it. A handful of brands showed up in nearly every answer, a long tail showed up in almost none, and the line between them had little to do with the SEO metrics most agencies still live by.
This is the reality of ai brand mentions in 2026: a parallel visibility layer, invisible in your current dashboards, where your clients are already winning or losing customers. Here's exactly how we ran the analysis, what the data showed, and what it means for the brands you manage.
How We Ran the Analysis
We selected 50 brands across five industries — B2B SaaS, direct-to-consumer e-commerce, financial services, professional services, and local/healthcare — ten brands per category, deliberately mixing category leaders with mid-market challengers so we could test whether size actually predicts AI visibility.
For each brand we built a set of roughly 15 to 20 realistic buyer prompts spanning the funnel: broad discovery questions ("what's the best project management tool for agencies?"), comparison queries ("is Brand A or Brand B better for X?"), and problem-first questions where no brand was named at all ("how do I stop losing leads from my website?"). That last category matters most, because it's where a brand either earns an unprompted recommendation or gets left out entirely.
Because large language models are non-deterministic — ask the same question twice, get two different answers — a single check proves nothing. So we ran every prompt multiple times across four engines: ChatGPT, Claude, Perplexity, and Gemini. For each brand we then measured four things: presence (did it appear at all?), share of voice (in what percentage of relevant responses?), prominence (was it named first or buried?), and citation (did the model link to the brand's own content as a source?). The entire analysis was run using AISurface.io, a free ai visibility tool that measures brand visibility across Claude, Perplexity, and Gemini, aggregating repeated runs into a stable visibility score rather than a one-off snapshot.
The result is a picture of how each brand actually surfaces in AI answers — not how it should based on its backlink profile, but how it genuinely does.
The Results: Winners and Losers
The single clearest finding: AI visibility is brutally concentrated. Across all 50 brands, the top ten captured roughly 68% of every brand mention we recorded. The bottom half split the remaining scraps.
Sixteen of the 50 brands — nearly a third — appeared in fewer than 10% of the responses to their own category's buyer questions. In practical terms, these brands are invisible. When a customer asks ChatGPT or Perplexity for a recommendation in their space, they simply don't come up. Not ranked low. Absent. And because AI answers don't have a "page two," there's no scroll that rescues them.
To give you concrete numbers: HubSpot scored 78% overall — Claude 83%, Perplexity 67%, Gemini 83%. Mailchimp came in at 22%. HSN, a major sports nutrition retailer, showed 56% visibility but with a striking gap: 83% on Claude and Perplexity, 0% on Gemini. At the bottom, a UK dental clinic and a Spanish marketing agency both scored 0% — completely absent from every AI answer in their category, while competitors like Yelp, Healthgrades, Ogilvy and Shackleton dominated the responses instead.
At the other end, nine brands appeared in more than 60% of their relevant queries, and a handful were near-ubiquitous — named in almost every answer, usually first, and frequently cited as a source. These are the brands compounding an advantage most of their competitors can't even see.
The most counterintuitive result was how weakly traditional strength predicted this. Several brands with the highest domain authority in their category landed in the invisible tier — strong on Google, effectively missing from AI. Meanwhile a number of smaller challengers, some with a fraction of the backlinks, dominated their category's AI answers. If you plotted domain authority against AI visibility, the correlation was loose enough to be nearly useless as a predictor. The old proxies for authority are not the new ones.
One more pattern: visibility traveled together across engines. A brand strong in ChatGPT was usually strong in Claude, Perplexity, and Gemini too. But the engines weighted their sources differently. Perplexity, which retrieves live, rewarded brands with fresh, linkable, recently-cited content. Gemini leaned heavily on what was well-indexed and authoritative in Google's ecosystem. ChatGPT and Claude leaned more on reputation baked into training data — meaning brands that had been widely discussed across the web for years held an edge that was hard to dislodge quickly.
What Separates High-Visibility Brands from Invisible Ones
When we looked at what the winners had in common, one factor towered over the rest: third-party validation.
High-visibility brands appeared, on average, across roughly three times as many independent sources as invisible ones — review platforms like G2 and Capterra, "best of" listicles, comparison sites, Reddit threads, industry roundups, and Wikipedia. This is the mechanism that trips up so many agencies. AI models don't primarily recommend a brand because of what that brand says on its own website. They recommend it because of what the rest of the web says about it. The invisible brands in our set often had beautiful, well-optimized sites — and almost no third-party footprint for the models to synthesize from.
The second differentiator was citable, extractable content. Winning brands consistently published material that answered buyer questions directly and cleanly: clear definitions, honest comparison tables, specific numbers, well-structured headings. Models pulling together an answer favor content that's easy to quote and hard to misread. Brands relying on vague, keyword-stuffed pages built to game older algorithms were routinely passed over, even when they ranked.
Third was factual consistency. Where a brand's pricing, product names, or positioning conflicted across its own site, its directory listings, and third-party sources, the models hedged, omitted it, or described it wrong. The brands that AI described confidently and correctly were the ones whose story was consistent everywhere the models could look.
Taken together, these three factors are the working definition of generative engine optimization — GEO — the practice of shaping how AI systems perceive and recommend a brand. The winners in our study weren't doing anything mystical. They were, deliberately or by accident, doing GEO: earning third-party mentions, publishing extractable content, and keeping their facts straight. This is what AI search optimization looks like in practice, and it's a meaningfully different skill set from ranking a page.
The Industries Where AI Visibility Matters Most
AI visibility isn't equally urgent across every category, and our five-industry split made the differences obvious.
B2B SaaS showed the highest stakes and the fiercest concentration. Software buyers have wholesale shifted to asking AI for shortlists — "best X for Y" is now an AI query far more than a Google one — and in our SaaS cohort the gap between the visible and the invisible was the widest of any industry. If your client sells software and isn't in the AI shortlist, they're not in the deal.
Professional services — agencies, consultancies, law and accounting firms — showed the most upside for challengers. Because these categories are under-documented by third-party sources compared to consumer brands, a firm that invests early in earning mentions and publishing citable expertise can leapfrog larger, older competitors quickly. The models are hungry for authoritative sources here and there simply aren't many.
E-commerce and DTC revealed a recency effect. Product recommendations shifted noticeably based on recent reviews and fresh coverage, and Perplexity in particular rewarded brands with current, linkable content. Visibility here is less "set and forget" and more of a moving target.
Financial services was the most citation-driven and the most cautious. Models leaned heavily on established, authoritative, well-regulated sources and were slow to recommend unfamiliar names — a high barrier to entry, but a durable moat for the brands that clear it.
Local and healthcare brands were the most invisible overall, largely because their third-party footprint tends to be thin and fragmented. That's a warning and an opportunity: the bar to stand out is low, and the brands that address it first will own their local AI answers.
What This Means for Your Clients
Here's the uncomfortable translation for anyone managing client relationships: your monthly report is very likely hiding this.
You can show a client a healthy rankings trend and a rising traffic line while their share of ai brand mentions quietly erodes — because those two things measure different worlds. Our data showed brands sitting comfortably on Google's page one while being nearly absent from the AI answers their customers now consult first. The client won't notice until a competitor is suddenly everywhere in ChatGPT, and at that point it will look like you missed a shift you had no instrument to detect.
That's the risk. The opportunity is the flip side of the same coin. Because AI visibility correlates so weakly with domain authority, it is genuinely winnable for smaller clients who could never outspend the category leader on links. A mid-market brand that commits to GEO — earning third-party mentions, publishing extractable content, tightening its factual footprint — can realistically outrank a giant in AI answers within a quarter or two. Our study is full of exactly those upsets.
And because most agencies aren't measuring this yet, you can walk into a review with insight nobody else is offering: here's precisely how you appear across the AI assistants your customers use, here's where you're losing to competitors, and here's the plan to fix it. That's a new, defensible service line built on a shift that's still early.
How to Check Your Own Brand's AI Visibility
You don't need to replicate a 50-brand study to get value from this. You need a baseline for the brands you actually manage.
Start with one high-intent question per client — the query they most want to win — and see how the major AI engines answer it today. That single result is usually enough to make the stakes undeniable: either the client's name appears, or a competitor's does. From there, build a proper prompt set around the client's real buyer questions, establish a baseline across engines, identify the queries where they're losing, and put a GEO plan against them. Then re-measure and show the movement.
That's exactly the workflow AISurface.io was built for. It's the same free ai visibility tool we used to run this analysis — it checks how your brand appears across Claude, Perplexity, and Gemini and returns a clear visibility read in about a minute, no audit required. Run your free scan at aisurface.io and see where your clients actually stand.
Conclusion
Fifty brands, four AI engines, and one conclusion that's hard to unsee: AI visibility is its own game, concentrated among the few who play it, and largely disconnected from the SEO metrics most agencies still report. The winners aren't the brands with the biggest backlink profiles. They're the brands the wider web talks about, describes consistently, and makes easy for a model to quote.
For agencies, that's the whole opportunity in a sentence. The shift to AI-assisted search is already reshaping who gets recommended, most competitors aren't measuring it yet, and the levers that move it — GEO, AI search optimization, a disciplined handle on ai brand mentions — are exactly the kind of work a good agency already knows how to do. The only thing missing has been the instrument to see it. Now that you can measure it, the brands that start first will be the ones showing up when it counts. Check your brand's AI visibility at aisurface.io.