AI visibility check
What the check measures
People now ask an assistant which product to buy before they search. The assistant answers with a short list, and a brand is either on it or not. This check asks that kind of question to ChatGPT, Claude, Perplexity and Gemini, each with its live web search switched on, and reads every answer for three things: whether your brand was named, which tracked competitors were named and in what order, and which sites the answer cited.
The questions are suggested from what we know about the company: its category, its country and its listed competitors. You can replace them with your own. Long, specific questions matter most for a smaller brand, because a broad question returns the five largest names and a narrow one returns whoever fits.
How the score is counted
A question that names your brand always gets your brand back. "Compare A and B" will mention A every time, and that says nothing about whether an assistant would bring A up unprompted. So those questions are shown for what the assistant says about you, and are left out of the mention count and share of voice. Only questions that do not name you are counted.
Position is counted among the brands being tracked: "Named #2" means one tracked competitor appeared in the answer before you. Answers change from run to run, so one run is a reading, not a ranking. The cited sites are the practical part: they are where the assistants got their list, and a site that is cited often and never mentions you is the place to get covered.
Free run and full check
The free run asks one question of one assistant, so you can see a real answer about your market. The full check asks every question of every assistant you pick, with your own questions, and is included in every plan. The same report is available through the API for use in your own dashboards and agents.
Related: AI Agent Readiness scores whether an assistant can read and use your site at all, and every free tool is in one place.
