Introducing the AI Visibility Index: We Asked 5 AI Engines Who Wins Secure Infrastructure Access, 50 Ways
A leaderboard for the AI answer layer
Today we're launching the Indexable AI Visibility Index — a recurring, quarterly Share-of-Model ranking of the brands that win AI answers, one B2B category at a time. Edition 1 covers secure infrastructure access, and it is built entirely on first-party data: 50 brand-neutral buyer questions, run across five AI engines (ChatGPT, Google Gemini, Perplexity, Google AI Overviews, and Google AI Mode) over roughly three weeks via Ahrefs Brand Radar.
The headline finding is that the engines don't agree. No brand exceeds ~24% Share of Model, and the #1 spot flips depending on which assistant a buyer opens.
Three things the data showed
- No consensus leader. Okta leads Share of Model at 24.3% (121 mentions), Teleport is second at 22.3% (111) — a two-point gap, with no brand dominant.
- Okta owns the chatbots; Teleport owns Google. Okta wins ChatGPT, Gemini, and Perplexity, while Teleport wins both Google AI Overviews and Google AI Mode. Same question, different "best."
- AI runs on video and community. youtube.com is the #1 cited source and reddit.com is #3 — ahead of every vendor blog. And being cited isn't the same as being recommended: Palo Alto Networks' domain was cited in 21 answers and recommended as a brand zero times.
The likely driver: Google's AI leans on the open web, where a vendor's own content can be one of the most-cited sources, while the chatbots lean on training data, where brand ubiquity dominates. That split is the whole reason AI visibility has to be measured per engine, not assumed from one.
Read Edition 1 in full
The complete leaderboard, the per-engine divergence, the cited-vs-recommended trap, the emerging challengers, and the full methodology and caveats.
This is Edition 1. Cloud Security / CNAPP, Endpoint / EDR, and Workforce IAM editions follow. If you want your own category — or your own brand's Share of Model — measured, talk to a strategist.