How does a Gap Audit measure AI visibility?
With your team, we define the six people on a typical buying committee, using your CRM to choose and weight them by how often each one decides or sponsors deals. We then write 60 questions those people would ask without knowing you: ten per persona, four at shortlist, three at evaluation and three at decision.
Every question goes to the five AI assistants your buyers use most, ChatGPT, Gemini, Perplexity, Claude and Google AI Mode, each with its own web search on and searching from your market. Each is asked as a first-time buyer would ask it: no login, no history. We run the full set three times, on days 1, 4 and 8, which gives 900 core answers and a margin of error of about ±3 points on headline measures.
Questions that name you, and three identity checks such as “What is [your company]?”, are reported separately. They show what known buyers are told, and they never count in the Gap.
- Up to ten tracked competitors, agreed with you, reviewed each quarter
- Each platform pinned to the model version most buyers get, recorded on every answer
- One market per Audit; another market can be measured as a separate, comparable run
6buyer personas, weighted by your CRM
60questions written blind, never naming you
5AI platforms: ChatGPT, Gemini, Perplexity, Claude, Google AI Mode
3runs over eight days
900core answers, each one open to read
±3points margin of error on headline measures