How I measure your visibility

Nothing magic, nothing opaque: I ask real questions to real engines, I read the answers, I count the mentions. Here is the method, step by step.

We build your questions

A customer doesn't type your business name: they describe their problem. “A good dentist open on Saturday”, “who can fix a heat pump urgently”. We build about thirty of these phrasings from your trade, your service area and your specialties.

That list is sent to you before the first measurement. You can remove, add or rephrase — it's your market.

We query the engines

Each question goes to Claude, Google AI Overviews and ChatGPT, in a neutral context — no history, no signed-in account, with a location matching your area. That's what makes two measurements comparable.

The raw answer is archived as-is. You can trace a score back to its source months later.

We separate visibility from reputation

Two distinct measures, often confused. Visibility answers “are you named?” and is counted. Reputation answers “how are you named?” and is qualified: warmly recommended, cited in passing, or mentioned with a caveat.

Being cited negatively is a different problem from being absent — and it isn't fixed the same way.

We fix what blocks you

A model cites you if it finds clear, consistent, corroborated information about you. The work covers three fronts: your pages, structured to be extracted; your listings, where name-address-phone consistency weighs heavily; your third-party sources, where others talk about you.

On Watch, you receive the list and apply it. On Growth and Dominance, we produce and deliver.

What we don't do

No fake reviews, no content farms, no hidden text injection, no prompt manipulation. These techniques work for a few weeks and then expose you to a delisting there's no coming back from. Everything we produce could be shown to your own customer without embarrassment — that's the only test we care about.

The first measurement is free

You'll see the method applied to your business before deciding anything.