How it works
People increasingly ask an AI assistant for a plumber, a dentist or a restaurant instead of scrolling a map. If the answer they get about you is wrong, you never hear about it — the customer just goes elsewhere. This closes that blind spot.
You add your business
Name, category and city are required. Hours and services are optional but matter more than they look — they are what we compare the AI's answer against, so without them we can flag that you are missing, but not that your hours are wrong.
We ask what your customers ask
Four questions every week: what the engine knows about you, whether you are any good, your hours and contact details, and who the best option in town is. That last one is how you find out you are invisible.
We flag what looks wrong
Claims that you have closed are critical. Wrong hours or denied services are high. Being absent from the "best of" answer is medium. We record the exact sentence that triggered each flag, so you can judge it yourself.
You get a report and an alert
A PDF lands in your inbox every Monday with scores, every question and answer in full, and the sources cited. Anything critical also fires an immediate separate alert rather than waiting for the weekly summary.
The two scores
- AI Visibility (0–100)
- Whether the engines can see you at all. 100 means you were named when asked for the best in your category. 70 is the baseline for being known but not recommended. 25 means the engine had no information about you, which is the quietest and most expensive failure mode there is.
- Accuracy (0–100)
- Starts at 100 and loses 30 for a critical error, 20 for high, 10 for medium, 5 for low. A single “permanently closed” claim drops you to 70 on its own, which is the intent.
What we are careful about
- An absence of data is not misinformation. If the engine says it has no information about you, we report exactly that. Dressing it up as a “finding” would mean inventing one.
- “Closed on Sundays” is not “closed down”. Opening-hours language is filtered out before anything gets flagged as critical, because an alert system that cries wolf is worse than no alert system.
- Every answer is stored in full. Not a summary, not our interpretation. You read what the engine said and decide for yourself.