AI visibility score: How often AI names you, and where in the answer.

Answers that named you, divided by the answers analysed in the window, as a whole percent. One answer is one prompt put to one assistant, so a prompt is counted once per assistant and again on every scan in the window. The same division runs for every brand you track, which is why the Rankings table and the trend can be read against each other.

It is the order of first mention. Verand finds where each tracked brand first appears in the answer text and numbers them from the top, so #1 means nobody was named before you. Average position is the mean over the answers that named you; a brand an assistant never names has no position at all, not a bad one.

ChatGPT, Google AI Overviews, Google AI Mode and Gemini run daily. Perplexity and Claude run weekly, on Mondays. That cadence is a property of the engine and is the same on every plan; what a plan changes is how many of the six you monitor. Run a scan by hand and it asks every engine your plan covers, whatever day it is.

Google often renders no AI Overview for a query at all. That is a real absence rather than a failure, since a brand cannot appear in an answer that was never generated, so it is not recorded as an error and no row is stored for it. Every rate on the page is counted over the answers that actually came back, never over the prompts you own.

Not by default. The Unbranded toggle is on when the tab opens and it drops any prompt whose text contains your name, because an answer to a question about you was always going to name you. Turn it off and those prompts are counted again, which is the reputation reading rather than the visibility one.

No. Nothing here writes to an assistant and no placement can be bought. What moves the rate is being the page an assistant reaches for, which is why this tab sits next to Sources, Topics and Mentions and Citations in AI Visibility: they name the domains being cited instead of you and the subjects where you are absent.