FAQ

Why does AI visibility matter? A growing share of buying research starts as a question to ChatGPT or Gemini instead of a Google search. If the assistants do not recommend you, you are invisible in those journeys — and unlike SEO, you cannot see it happening without measuring.

Do you scrape ChatGPT? No. Every provider is queried through its official API with the consumer-default model of its assistant. Clean, reliable measurement — no scraping, no terms-of-service games.

Why does my score differ between plain and grounded mode? They measure different things: what the model learned in training versus what today’s web search returns. Grounded mode is part of the paid plans. See Providers, modes and languages.

Why did the index jump when I added a language or a question? The index averages every run; changing the tracked set changes the denominator. Compare per-slice trends after a change, or keep the set stable for a clean blended trend.

Can I influence what the assistants recommend? Yes — that is the point. Answers draw on training data and, in grounded mode, on the sites they cite. Citations and the advisor show which sources shape your market and what to publish.

How do aliases work? The brand profile lists the names the assistants might use (product name, company name, common misspellings). Mentions are matched against them; competitors have their own alias lists.

What happens on the free plan? One project, five tracked questions, one language, five competitors, weekly checks in plain (model-knowledge) mode, alerts and the weekly digest — free forever, no card. Live web-search mode and the advisor come with the paid plans. Plans.

Can I export the data? The public report has a print/PDF view, and every score is visible per day in the dashboard. There is no CSV or API export today.