The audit runs free, on your site or ours
The technical audit is public and reads only pages that are already public. Point it at your own domain, or at katman.pro, and compare what it reports to what we claim on these pages.
Run the audit →katman / What is katman
RNIFY LIMITED · ENGLAND · FORMED 4 SEPTEMBER 2026
katman is an AI visibility measurement service. We put the questions your buyers actually ask to five model families — OpenAI, Anthropic, Google, xAI and Perplexity — with live web search, record every answer verbatim with its sources and exact model version, and turn what is missing into specific work you can publish.
A company that sells measurement should not bury its own. Three things about us are fair to hold against us, so they go at the top rather than in a footnote:
The only honest response to all three is to make every claim checkable without needing to trust the person making it. That is what the next section is.
The technical audit is public and reads only pages that are already public. Point it at your own domain, or at katman.pro, and compare what it reports to what we claim on these pages.
Run the audit →Thirty observations of frutti.ai across five model families and four languages. The brand appeared in none of them. Every answer is reproduced verbatim with its sources and model version.
frutti.ai run, 4 September 2026 →Twenty-eight days of one site's sessions and revenue by channel, including the four attribution mistakes we had to fix before the numbers meant anything.
AI traffic study →102 pages across the six sites AI assistants recommend in this category, measured against our own site — which lost on most of the measures.
What is GEO, with the data →We run the audit against katman.pro and publish what it finds. At the last run it reported zero critical issues and zero of seven buyer queries matched by a page title. The second number is why this page exists.
Observation log →An agency sells retained hours. We sell a dated measurement and the written work that follows. Nobody here posts or publishes anything as you.
There is no rank to track. A model either has a usable sentence about you or writes the answer without you in it. We report appearance across a question set, not a position.
Nobody can sell placement in an organic AI answer. Anyone offering one is selling something else, and we say so on every page where it could matter.
katman is the trading name. The contracting party is Rnify Limited, a company registered in England, formed on 4 September 2026. Invoices, contracts and any refund under the Verified Program are issued by that company. If you need the registration details before buying, ask on the contact page and they will be sent to you directly.
Companies that already sell something buyers compare before purchasing, and that suspect the comparison is now happening inside an assistant rather than on a results page. It is most useful when there is a real category — software, tools, services people ask for by problem — and least useful for a brand nobody would ever ask a model to recommend.
The technical audit is free, public, needs no card and reads only pages that are already public. A one-time baseline measurement is $299. The three-month Verified Program is $999, and on that one 90% of what you paid comes back if mention and recommendation rates have not moved by day 90. Full pricing and what each includes.
katman is an AI visibility measurement service. We put the questions your buyers actually ask to five model families with live web search, record every answer verbatim with its sources and exact model version, and turn what is missing into specific work you can publish.
katman is operated by Rnify Limited, a company registered in England. We do not currently publish the names of the individuals behind it. That is a fair thing to hold against a company selling measurement, and the only useful answer is that nothing we claim depends on trusting us: the audit is free and runs against any site including ours, every measurement run is published in full including the ones that embarrass us, and each report carries the exact model version and timestamp so you can repeat the run yourself.
Rnify Limited was formed on 4 September 2026 and katman is what it operates. It is a new company and there is no way to present that as a track record. What exists instead is published evidence — a 30-observation model panel, a first-party traffic study and a 102-page competitor crawl, all dated and reproducible.
England. The operating company is Rnify Limited, contracts and invoices are issued from the United Kingdom, and the service is sold worldwide with no geographic restriction on who can buy it.
No. An agency sells retained hours. We sell a dated measurement and the written work that follows from it. Nobody here posts, publishes or edits anything as you — we write what needs publishing and you decide whether to ship it.
Both, and the split is deliberate. The technical audit is a free, public tool that reads only your public pages. The model panel is a paid service because every run spends real provider credits with OpenAI, Anthropic, Google, xAI and Perplexity.
You should not, on trust. Check the artifacts instead: run the free audit against your own site and against ours, read the published frutti.ai run where a brand appears in zero of thirty answers, and repeat any single question from a report in your own signed-out chat. If a claim cannot survive that, it should not have been made.
We do not guarantee a position in any AI answer, do not report a single score out of 100, do not count questions containing your own brand name toward visibility, and do not publish anything on your behalf.
Check us the way we would check you. Run the free audit against your own domain, then read a published run in full before deciding whether any of this is worth paying for.