GapCheck / the entry reading
Who gets named when your buyers ask AI about your category?
GapCheck is free: we run your buyer prompts against AI engines and report who the answers name, where they source from, and whether you exist in them. A human verifies every report before it ships. 24-hour turnaround. Your reading stays private: we never publish a report about a company in our pipeline; published teardowns are our own properties or labeled category readings.
The variance-honesty promise
We never report a single fake-precise AI rank. Every result in a GapCheck report shows its run-to-run band and per-engine disagreement, and single-run readings are labeled directional only.
Any tool giving you one number is selling noise.
Request a report
Is there a free tool to see which brands AI answers mention?
Yes. GapCheck is free: it shows which brands ChatGPT, Perplexity, and Google's AI name for your category. A human verifies each report in 24 hours.
It costs your work email, nothing else. You give it your domain and category; a human runs your buyer prompts through the engines, checks the readings, and sends back who the answers name, the pages they cite, and the run-to-run variance. See the format on our own category teardown.
How do I check my website's AI search visibility?
Run your buyer prompts through the AI engines and see which pages they cite. GapCheck does that for you, free and human-verified, in 24 hours.
Per prompt you get an absence-or-presence verdict, the grounding domains behind the answer, and the variance band, so you can tell a real gap from run-to-run noise. That is the reading a daily rank number cannot give you.
What does the GapCheck report contain?
Per buyer prompt: which brands the AI names, your absence or presence verdict, the cited grounding domains, and the run-to-run variance band.
Plus the basket-level share of voice across your category prompts and a grounding concentration read per prompt, because a prompt whose answers cite 4 domains is winnable in a way one citing 40 is not. See the format on the live sample: the teardown of our own category, absence verdicts and all.
Why is the report human-verified instead of instant?
A human runs and checks every reading before it ships. Instant scanners hand you one noisy run; we would rather be right than instant.
An automated scan that misreads your category or your brand string sends you a wrong verdict with a confident face. We check prompt selection, brand extraction, and the verdicts by hand before anything leaves. That is a positioning choice, and it holds until volume forces the self-serve version.
Will GapCheck give me my AI rank?
No. Single-query AI rank is statistical noise: under 1 percent repeatable in the SparkToro/Gumshoe study. We report basket-level share with its band.
If you want a dashboard with a daily rank number, several tools sell one. What they do not show you is how much that number moves when you run the same query twice. We show you exactly that, because the decision you make on top of the data is only as good as the data's honesty about itself.
What does GapCheck cost?
The GapCheck report costs your work email. The deeper read is $750 for inbound requests, credited in full toward the roadmap or first retainer month.
Why "for inbound requests": when we reach out to you, the first read is on us; when you ask us to work on spec, it is $750. The paid read filters for seriousness and pays for the human verification time. It credits against the roadmap or the first retainer month. Terms on the services page.