How we condense five LLMs into a single number, why we exclude informational prompts, and what a good score looks like in your industry.
What the Index measures
The Visibility Index combines two signals per LLM. The first is the mention rate: how often your brand name actually shows up inside the answer text. The second is the link rate: how often the model cites your domain as a source. Both matter, but they matter differently. The brand mention is what a human reads and acts on. The source citation is the conversion mechanic afterwards. We weight them accordingly, with mentions counting noticeably more than links.
Once those two signals are blended for each LLM, the per-model scores get rolled up into one combined number on a 0 to 100 scale. The aggregation is not a flat average. Each model contributes proportionally to its real-world usage in your market. A brand that does well only in a niche model and is invisible in the model your buyers actually open should not score the same as a brand that wins the mainstream answer engine.
Why ChatGPT carries the most weight today
By every measurement that exists in mid 2026 (Similarweb traffic data, the a16z AI consumer report, Pew's annual survey, Statista's panel work), ChatGPT is still the most-used consumer answer engine by a wide margin. Gemini comes next, lifted by deep integration into Android, Workspace and Google Search itself. Claude has the smallest user base of the major players but the highest per-user engagement in B2B and developer contexts. Perplexity and Grok round out the spread, smaller but visibly growing.
The combined score reflects that distribution. When Gemini took share off ChatGPT through late 2025 the weighting moved with it, and we retune the distribution at least twice a year. You can always see your per-LLM breakdown in the dashboard. The headline number on the dashboard tile is the market-weighted view, so you read one figure and know roughly how visible you are in the AI answer flow buyers actually use.
Why we exclude informational prompts
"What is X?" doesn't drive purchase decisions. Tracking it inflates your visibility number with traffic that never converts. We filter pure-informational prompts from the index automatically so the score reflects buying intent, not curiosity.
What's a good score
Industry medians from our customer base (May 2026):
- SaaS B2B: 28 to 45 is healthy, 50+ means you're a category leader.
- Hospitality: 35 to 60 is normal, big chains 65+.
- E-commerce: 20 to 40 typical for established DTC, 60+ for category-defining brands.
Your number alone tells you little. The trend over time, the per-LLM breakdown, and the gap to your top 3 competitors are what matter.
