Who is this for, and who is it not for?
For a business whose customers come from a few kilometres around: a hairdresser, a garage, a dental practice with nobody on marketing, a restaurant. What they share is not the sector but the question — someone looks for something nearby and gets one recommendation.
| Who | Who decides | What keeps them up at night |
|---|---|---|
| Self-employed with a local customer base | The owner, usually in the evening | What it costs, and whether this is one more thing on the pile |
| Hospitality | The operator | No time, and the weight of reviews that stay up |
What weighs in when an AI picks one place nearby?
Four things, and all four are fixable. The right-hand column is there because we would rather say where the measurement stops than pretend it goes everywhere.
| Driver | What ceeme measures underneath | What stays dark without a connection |
|---|---|---|
| Local trust | GBP, reviews and whether your details match everywhere | We read your Google profile. Changing anything in it needs a GBP connection. |
| Reputation risk | misquotes, sliding reviews and where they come from | Reviews and social are read live. Print and paywalled media are not in there. |
| Regional relevance | whether you score the same in one region as in the next | For Belgium this is exact. For a region without its own registers it is an estimate, and it says so. |
| Channel readiness | whether your profiles are complete and alive | Measuring always works. Putting something on that channel needs a ProOps connector. |
| Design trust | whether your site earns trust from someone who hesitates | We see what is wrong on your site. Fixing it goes through a CMS connection. |
Where does an AI assistant get its picture of my business?
Almost never from your website alone. An engine answering "a good hairdresser near me" leans on your Google profile, on review sites and on local guides — places where you do not publish. That is why a beautiful new site sometimes changes nothing about the answer.
| Source | Kind |
|---|---|
| GBP/Maps | public — we read it, you connect nothing |
| reviews and directories | public — we read it, you connect nothing |
| public — we read it, you connect nothing | |
| public — we read it, you connect nothing | |
| TripAdvisor-style guides | public — we read it, you connect nothing |
| GBP (read) | connected — only after your approval, read-only unless agreed otherwise |
| GA4 (optional) | connected — only after your approval, read-only unless agreed otherwise |
Can I use this to boost my reviews?
No, and that is not caution but a hard rule in the code. Fake reviews and bought local signals do not come out of here, not even on request. They also only work briefly: an engine that reads a review pattern as suspicious downgrades the whole profile.
| Rule | Kind |
|---|---|
| We do not write, buy or provoke reviews, and we do not manufacture local signals that do not exist. | forbidden — we do not do this, not even on request |
| No review that appears on your pages comes from us or from a service we hire. | forbidden — we do not do this, not even on request |
What do I need to have ready, and what does it cost me in time?
Read access to your Google Business profile, your website address, and ten minutes to say which questions your customers ask. That last one is the only thing we cannot guess, and it decides the whole measurement.
Frequently asked questions
I already rank top on Google Maps. Do I still need this?
It helps, but it is not the same thing. Google Maps shows a list and you are somewhere in it; an AI assistant gives one name and the rest does not exist. Whoever ranks third on Maps but goes unmentioned in the AI answer loses a customer who never looked at a map. That gap is exactly what Local Proof measures.
Do you carry out the changes yourselves?
Not in this pack. Local Proof prepares: you get the exact change with the source that makes it necessary, and you apply it. If you want us to do it, that is a ProOps route, with a connection and approval per action. That boundary is not in the fine print but in the code.
How do I know whether anything improved?
By asking the same questions again, with the same method. Not by a number that goes up without anyone being able to say from what. If nothing changes, that is written down too — a re-measurement that only counts the wins measures nothing.
Read on
Free, no account and no card.