Six things have to happen in order, every day, for a visibility number to mean anything. Citino is those six steps wired together. The five feature pages below are the five places you look; this page is the machine they belong to.
Every morning the same questions are asked, read, measured, explained, priced in sessions and checked against what you changed. Hover a step to stop the loop on it.
The loop runs whether or not you are logged in. The product is the place you read it.
A category is defined by its 54 prompts, written as the questions buyers actually type — "best warehouse management software for a 3PL" rather than "WMS vendors". They are asked daily against ChatGPT, Gemini and Perplexity on a fixed schedule, times in ET. Claude is available as an add-on.
The wording is then immutable. A prompt cannot be edited, only retired and replaced, because a trend line across two different questions is not a trend line. Copilot and Google AI Overviews are not sampled at all — neither has an official answer API.
Parsing an answer means deciding that "Northport" and "Northport WMS" are one company, that a brand named in a caveat is not a brand recommended, and that the third name in a list is third. Citino records the mention, the naming order, the tone of the sentence, and every URL the engine cited.
Every answer is archived raw. When the matching logic gets better it re-reads the whole archive and the history improves with it. A tool that stores only the extracted numbers can never correct a past mistake, only stop making new ones.
Every answer, stored whole and searchable. → Answers
Four numbers, scoped by segment, each printed with the sample it came from and a 95% interval. A change inside that interval is labelled within noise and does not alert. Tallyfold is up 1.7 on a band of ±3.0, which is not a result yet.
Ninth of 17 brands in Warehouse Management Software · US. → Measure
Visibility without a cause is a scoreboard. Citino collects every URL cited across the category's answers and ranks them by leverage: pages you own sit in front of pages you can influence, which sit in front of pages you can only watch.
The engines lean heavily on third-party comparisons, listicles and forum threads, and seeing which specific ones they cite is usually the difference between a quarter of guessing and an afternoon of work.
Cited across 1,785 answers in 30 days. → Sources
Citino joins the answers to your own GA4 traffic and matches AI-referred sessions back to the prompts that plausibly produced them. That 29% is printed, not distributed — the honest models show you how big the residue is rather than spreading it so the pie adds to 100%.
The top prompt by value — "best warehouse management software for a 3PL" — carried 186 sessions and 9 demo requests.
Copilot and Google AI Overviews appear here, and only here, as labelled referral sources. They are not sampled. → Attribution
You rewrite the 3PL billing page. Visibility on the affected prompts goes 21.1% → 29.5%. Over the same window the whole category rose 0.5 points without doing anything, so the net effect of your work is +7.8 points. It will not be called a win until the delta clears its band.
Citino does not run a collection per customer. It runs a collection per category, and measures every brand in it whether or not any of them is a customer. One index, every subscriber.
So the price does not rise when you add competitors to watch, and the benchmark pages can be public, because they are real measurements of brands who never signed up. A per-customer tracker cannot match this by adding a feature — it would have to change what it collects and who it collects for, which is a different company with a different cost base.
Only custom prompts scale per customer. Those are the ones nobody else can share, and they are the only line on the bill that grows.
Pick a category and the six steps start on it. If it is already live, the first screen you see is the league table.