We measure the answer, not the ranking

For twenty years, being found meant ranking on a results page, and an entire industry grew up around measuring that. Then people started asking assistants instead. An assistant does not hand over ten blue links. It gives one answer, names two or three businesses, and moves on. If you are not in the answer, you are not in the conversation.

When we went looking for a way to measure this for an ordinary business, what we found was guesswork: someone types three questions into ChatGPT and screenshots the result. That is not a measurement. It is one sample from a system that answers differently every time you ask.

So we built the boring version instead. Thirty questions, four models, three runs each, every answer stored, every claim checked, every source traced. What you get is a number you can compare month to month, and a short list of things to fix that moves it.

Real questions, not keywords

Buyers ask assistants in full sentences with context. Our question sets are written the same way, and we never include your business name in them.

Show the evidence

Every metric in a report links to the quoted model responses behind it. If you disagree with a flagged error, you can read exactly what the model said.

Fix the source, not the symptom

You cannot edit an AI model. You can edit the pages and listings it learns from. Our recommendations focus on those.

Don't oversell the channel

AI recommendation is one channel among several. We tell you what it is worth for your business, including when it is not yet worth much.