Free tool, no sign-up
AI Agency Claims Checker
Bring the claim you want to check.
Bring the result and the evidence behind it. We can talk through what still needs checking.
Behind the tool
Why verification questions beat gut feel
Most AI agency claims are unfalsifiable as written: no date, no definition, no artefact. The questions this tool generates force each claim into one of two shapes: a number with evidence behind it, or an adjective wearing a number's clothes. Our guide to verifying agency claims walks the method end to end.
Questions
Frequently asked questions
How does the claims checker work?
It classifies the claim you paste (ROI figures, client counts, testimonials, guarantees, speed promises) with a transparent rule table, then shows the verification questions a careful buyer would ask for that claim type, plus what real evidence for it would look like. There is no AI model behind it: the rules are fixed, instant, and the same for everyone.
Why would an AI agency publish a checker that works on its own claims?
Because our positioning depends on it. Every number we publish sits in a public claims registry with the command that re-derives it, and our case studies disclose their nature. Run our claims through the checker; the questions it asks are the ones we already answer.
What is the most common red flag?
Numbers that drift: the same metric appearing with different values on different pages of the same site. It usually means the counters are hand-maintained marketing copy rather than the output of any measurement system.
What should I do with the answers?
Treat them as a scorecard. An agency that can answer most questions with dated artefacts is safe to shortlist. One that answers with adjectives is telling you how the engagement will go.
Keep exploring