When AI chatbots recommend a roofer, the names they give often differ from the businesses leading Google Maps. A new study found that about two in three roofers named in AI answers sat outside their city's Google Maps top 10, a gap that matters for any local business counting on AI visibility.
The study covered 20 US cities on October 15 and 16, 2026, asking ChatGPT, Perplexity, Gemini, and Google AI Overviews the same roofer question in each city, for 80 queries in total. Holding the question constant across cities and platforms is what makes the comparison meaningful: every engine faced the same task, so differences in the answers say something about the engines rather than about how the question was phrased.
The results were then checked against Google Maps rankings for each city. That choice of benchmark is deliberate, because Maps is where many homeowners actually start when they need a roofer urgently. A business can be well reviewed and still sit outside the Maps top 10, and the study set out to measure how often AI recommendations landed in that position.
AI answers named 199 rated roofers, of which 133 sat outside their city's Google Maps top 10. That is about two in three, and it is the figure local SEOs should remember first. It does not mean the AI-named businesses were poor choices, since every one of them was rated, but it does mean that AI visibility and Maps visibility are clearly different things, and winning one does not guarantee the other.
The figures cover one two-day sample, not a permanent ranking, and both AI answers and Maps results can shift over time. Still, the size of the gap across 80 queries suggests it is not a quirk of a single city or a single platform. A roofer strategy built only around Maps, or only around AI answers, would miss a large share of the other channel.
Roofers named in AI answers had a median of 289 Google reviews, against 242 for the Google Maps top 10. That difference cuts against any simple assumption that AI picks weaker businesses. On review volume at least, the AI-named roofers looked at least as established as the Maps leaders, which suggests the engines are weighing signals beyond the Maps ranking itself when they choose whom to name.
For roofers, the lesson is to treat review volume as an asset that travels across channels. Reviews earned for Maps visibility may also be helping a business get named in AI answers, even when its Maps position alone would not have put it in front of that customer. Building a deep, genuine review base remains one of the few investments that appears to pay off in both places.
Perplexity cited a review or directory site in 18 of its 20 answers, AI Overviews in 6 of 7, ChatGPT in 6 of 20, and Gemini in 2 of 20. That spread is at least as important as the ranking gap, because a citation is how a customer checks a recommendation and how a business earns a visit. An answer that names a roofer without pointing anywhere leaves the customer to search again, while an answer with a cited source gives the business a direct path to being found.
The pattern also hints at different roles for different engines in a local search journey. A heavily cited answer works more like a research shortlist, while an uncited answer works more like a word-of-mouth suggestion. Roofers who want to be cited need to be present and well described on the review and directory sites these engines actually draw on, not just on their own websites.
The practical takeaway is to track AI answers and Maps rankings as separate channels with separate scoreboards. A business in the Maps top 10 should still check whether AI engines name it, and a business named by AI engines should still check its Maps position, because the study shows the two lists overlap far less than many would assume. Reporting one without the other gives an incomplete picture of local visibility.
None of this replaces the basics: accurate business information, strong reviews, and a website that answers the questions homeowners actually ask. What it adds is a new checkpoint. Before assuming a local strategy is working, ask the engines themselves the question a customer would ask, and see whose names come back.
Sources: Overtake Marketing via EINPresswire, October 30, 2026.
Linkbaits helps you create content that earns citations and backlinks. Start with the free tools and build your first linkable asset this week.
Start free trial