This article was sponsored by Victorious. The opinions expressed in this article are those of the sponsor.
When we ask an AI platform to describe a brand, it almost always knows the answer. But when a buyer asks this same platform which brand to consider, they only rely on a small fraction of the brands they know.
This gap is the central finding of our Q2 2026 quarterly research report. Across eight AI platforms, 96% of the brands we tested were accurately described when we asked about them directly. Again 89 percent of the brands we measured never appear in AI-generated answers to category search questions.
AI systems already know most brands. Getting mentioned requires more than just brand recognition. This study aimed to measure what other factors influence the visibility of AI.
The Study Methodology: Measuring Recognition of AI Against Mentions of AI
We measured two distinct signals for a cohort of 175 brands across five verticals: legal, healthcare, SaaS, financial services and e-commerce/retail. Among these, 140 brands had AI-evaluable responses for the recognition analysis, and 150 formed the mention rate cohort.
Measuring AI Recognition
We asked eight AI platforms (ChatGPT, Claude, Gemini, Copilot, Perplexity, Google AI Overviews, Google AI Mode, and Meta AI) to describe each brand in our cohort, then evaluated each response against the brand’s website and assigned that response a label of correct, wave, exceeds, fakeOr not recognized.
Measuring mentions of AI
We measured how often these brands appeared in AI-generated responses to standardized category search prompts (the shopping search questions a prospect asks when comparing solutions). We also ran a separate set of problem awareness prompts (the questions asked early in a buyer’s journey before knowing what type of solution they need), which became important later in the study.
Broaden the analysis
The recognition and mention rates were not what we expectedso we extended the analysis to organic rankings, organic traffic, third-party web mentions, and Knowledge Graph presence.
AI recognizes many more brands than it mentions
Recognition and mention are distinct and measurable behaviors. Recognition reflects the AI’s understanding of a brand. The mention measures whether AI platforms would name a brand while a shopper compares options.
When asked directly, AI platforms accurately described 96% of brands. They correctly explained what each company sells and the markets they serve. Yet these same brands were rarely mentioned when we asked the category research questions that a real shopper would ask. For 89 percent of them, they never surfaced at all.
Upon closer inspection, we found that recognition varied by platform. Google AI Mode, Gemini, ChatGPT, Google AI Overviews, and Copilot all exceeded 83% accuracy across all verticals we tested, while Perplexity and Meta AI stretched much further. Perplexity correctly recognized less than 55% of SaaS and e-commerce brands, and Meta AI only recognized 46% of SaaS brands. The full report shows which platforms recognize brands most accurately in each of the five verticals.
So what separates the brands mentioned by the AI?
If recognition doesn’t explain the difference, something else does. The strongest relationships we found came from signals beyond a brand’s own website: referring domains (how many unique sites link to you) and third-party web mentions (how often other sites talk about you). Both showed a moderate correlation with mentions of AI, 0.49 and 0.45 on a scale where 1.0 would be a perfect relationship. No signal is powerful enough to act as a lever on its own, but together they describe a brand’s overall awareness across the web.
After edition This report, We Also tested if referring domain quality mattered as a lot as quantity. He is it that‘t. High authority links added only A modest elevator on their own, And a few brands with links Since very authoritative Sites always rarely appeared In AI answers. We to break down why these high authority links didn’t help on A recent episode of OUR podcast.
Third-party mention data was the most practical benchmark in the report. Brands with fewer than 2,000 indexed web pages mentioning them were only named in the AI responses 3% of the time. The mention rate climbed sharply from there, and the full report is down the visibility criteria which separated the brands mentioned at each threshold.
AI trusts different sources at every stage of the buying journey
This is where the data provided more actionable insights. In our category search prompts, 99.99% of the 49,391 citations we analyzed pointed to third-party websites rather than the brand’s own domain. Only 4 of 150 brands in our mention cohort earned a citation on their own website. This raised a follow-up question: Does the same pattern hold earlier in the journey, before a buyer starts comparing solutions? We compared the sources cited by AI in our two types of prompts to find out.
Issue awareness questions, such as “How do I know if my business needs a lawyer?” » were cited from educational sources: videos, government and institutional sites and professional communities. They also cited educational content posted on company websites. Yet the brands themselves were only mentioned in 0.10% of these responses.
Matching matters more than either metric alone: At the start of the journey, the AI will use a brand’s educational content to craft a response while leaving out the brand name. Just as recognition does not guarantee a mention, neither does a citation.
Category search questions, such as “What are the best law firms in the United States?” “, moved the citations to directories and comparison sites. And brands were cited more than twelve times more often in these responses.
The two stages reward different work. Post educational content that answers the initial questions in the journey, because that’s what the AI is looking for to create responses to issue awareness prompts. And build your presence on directories and comparison sites that AI visits when searching for categories, because that’s where brands are actually named. The report details which sources AI trusts at each stage of the buying journeybroken down by prompt type and platform.
What else did the data show
The sources on which AI relies also differ greatly depending on the sector. Legal services citations were concentrated in a handful of prestigious directories, while SaaS citations were scattered across more than 10,000 unique domains. So, the right offsite strategy highly depends on your vertical. The full report breaks down how citation sources differ by industryincluding the most cited sources for each of the five verticals we studied.
Key to remember
The AI already knows who you are. He describes most brands accurately and has no trouble explaining what they do. Visibility comes from appearing in the sources AI trusts at every moment of the buying journey. Your own educational content can earn citations while buyers are still working to understand their problem, and third-party sources carry almost all the weight once they reach the review stage.
The full quarterly research report for Q2 2026 includes full signal correlation data, third-party mention thresholds, citation behavior by platform and market signals this quarter.
Image credits
Featured Image: Image from VictoriousSEO. Used with permission.
In-Post Images: Images from VictoriousSEO. Used with permission.
