A Boise-area roofer (sample)

What we scanned. For each town listed in this report, we ran live Google Maps searches for the search terms shown, from a neutral location profile, on the date stamped at the top of this report. Your position is where your business actually appeared in those results. "Who gets the call" shows the businesses that held the top spots in that town at scan time.

Which searches, and why. The search terms are the ones a customer types when they are ready to hire: a named service, or an urgent problem with "near me" or the town attached. Research questions such as "how much does a new roof cost" are a different kind of search that the map pack does not answer, so they are kept out of the grid unless a report asks for them as a separate layer. Mixing the two would make a business look absent on a search it was never going to appear for.

How the comparison is built. Every competitor in the grid is a matched control: the same search, the same town, the same neutral device profile, on the same date. When you did not hold a top-three spot, the row shows who did on that exact search, so the gap is between two listings under identical conditions, not between your listing and a general top-ten list.

How the AI answers were checked. We asked ChatGPT, Claude, Gemini, and Perplexity the same question a homeowner would ask, three separate times each with live web search, on the date shown. We report how many of those runs named your business and how many named each competitor. We show counts, not single answers, because AI assistants do not give the same answer every time.

The control band. Three runs per assistant give a frequency, not a verdict, so we do not call a business absent from a bare zero. The businesses the assistants named in at least half of the runs are the control set. The fewest runs any of them was named in is the floor, and the runs they went un-named in are the band of absence that happens by chance for a business the assistants do recommend. Your count is read against that band: below it and never named is a real gap; at or above it is within normal variance. If no business cleared the control threshold, the answers were too inconsistent to call anyone absent, and the report says so.

Why results can vary. Google Maps rankings shift with location, time, and competition. AI answers vary between sessions, phrasings, and accounts. A single search on your own phone may differ from what this report shows. What matters for your phone ringing is the pattern across many searches by many customers, and a pattern is what this report measures.

What we did not do. No business listed here was contacted, endorsed, or reviewed by us. Competitor figures (ratings, review counts) come directly from public Google data at scan time. Every figure here is measured, except the modeled monthly-dollar range, whose method is shown with it. We keep the raw data for every scan in this report on file. Tool mentions are informational only, not paid endorsements or affiliate placements.

Questions about a number? Reply to the delivery email. If any figure in this report is wrong, we will correct it and extend your scan at no charge.