How the snapshot is actually run
The method, in public
A single AI answer is not a stable fact. It changes with the session, the time of day and the person asking. So we do not report one answer — we run a fixed set of questions several times and report exactly what we saw, with timestamps.
The questions we ask
The same question templates are used for every customer, so results are comparable between businesses and over time. The placeholders in brackets are filled in with your trade, city and area before we run them, and the exact wording we used for you is reproduced in your report.
- Who is a good plumber near [AREA]?
- Recommend an emergency electrician in [CITY]
- Which HVAC companies in [CITY] have the best reviews?
- I need a plumber in [CITY] this week — who should I call?
- Best rated [TRADE] in [CITY]
How we run them
- The same question set is run in at least three separate sessions, at different times of day.
- Each session is started fresh, with no prior conversation and no memory of earlier runs.
- We run them in ChatGPT, Gemini and Google AI Overviews.
- We record the date, the time and the assistant for every run.
- We record every business named in every run — not only whether you appeared.
How we report the result
The scoring is plain, with no interpretation added beyond what the runs actually showed:
- If your business is named in every run, we say so.
- If it is named in some runs and not others, we report it as a fraction — for example, "named in 2 of 5 runs on [DATES]" — and the report explains what that means: inconsistent presence, which is usually worth fixing with the same changes as outright absence.
- If it is never named, we report exactly which businesses were named instead, so you can see who is currently taking the spot.
We never write "the AI does not recommend you." We only report what we saw, and when we saw it.
What can make your own check disagree with ours
- Different phrasing of the question than the templates above.
- A different location signal (your phone's location versus a typed city name).
- Being logged in versus logged out of the assistant.
- A different version of the assistant than the one we used.
- Simple run-to-run variation — the same question can get a different answer minutes apart.
If your own check contradicts our snapshot within 7 days, see the refund policy.
What we do not do
- We do not guarantee inclusion in any AI assistant's answer.
- We have no relationship with OpenAI, Google or any assistant maker.
- We do not write or steer your reviews.
- We do not arrange or encourage incentivised reviews.
Ready to see your own nine runs?
Order the snapshot — 49 EUROne-time payment. PDF by email within 48 hours.