How we check

We start from the people who would be asking: a patient looking for a clinic near home, a family relocating and looking for a broker, a founder looking for an accountant. We write questions the way they would type or say them, including area names and specific services.

A typical scan uses around 40 questions for a category, grouped by service. Individual checks use questions about the person by name, and questions about their profession in their city.

ChatGPT with search turned on, Google Gemini, and the AI Overviews shown at the top of Google search results. Each question is asked in a fresh session with no account history, from a location in the city being checked.

The same question can get a different answer a minute later. We ask each question three times on each engine and count across all runs. A business that appears in one run out of three is treated differently from one that appears in all three.

A business or person counts as named if the answer mentions them by name as an option or recommendation. A passing mention in a list of sources does not count. We record every source the answer cites or links to.

The AI Mention Rate is the number of answers that named you divided by the total number of answers we checked.

For the Google comparison we run the same search on Google and record the order of businesses in the map results, from the same location. For individuals, we record the top results for a search of their name.

Results are a snapshot of a few days. AI tools change often, and answers vary with wording, location and whether someone is logged in. We can’t see inside the AI systems, so we report what they said and which sources they showed, not why. The bands are our own and are based on the scans we have run so far.

A report is a snapshot, not a verdict.