Only 7% of the businesses AI assistants named first were in the Google Maps top 3 for the same search. 68% were not in the Maps top 20 at all. That is from a study that asked ChatGPT, Perplexity, Gemini, and Google’s own AI Mode to name the best local business 1,600 times, then compared the answers with Google Maps rankings.
Here is what that looks like in practice. In Columbus, Ohio, ask Google Maps for an electrician from downtown and the top three listings have 39, 3, and 2 Google reviews. Ask all four AI assistants who the best electrician is, and they all name the same business first: one with 2,967 reviews that Maps puts at number 19. Same city, same trade, same morning. Two completely different answers.
The study, run by Awkward Media in November 2026, is one of the first to systematically compare AI recommendations with Maps rankings. If you own a local business, the gap between where you rank on Maps and whether AI names you matters more than you might think.
How Often AI and Maps Disagree
The numbers are striking. Across 200 searches in 40 cities and 10 trades, only 7% of AI first picks were in the Maps top 3. In 102 of 200 searches, no two of the four AI assistants named the same business first. A day later, the first business named had changed in 58% of cases.
A separate study by Nimza, covering 603 cities and 33 trades, found that the number one business on Google Maps was named by any AI engine in only 87 of 594 cities, or 15%. Of the 1,752 businesses in the Maps top three across every city, only 12% were named by at least one AI engine. 2,750 different businesses were named across 3,558 recommendation slots, and 2,172 of them were named by only one engine.
What this tells you: ranking well on Google Maps does not guarantee AI assistants will recommend you. And being recommended by one AI assistant does not mean the others will agree. The AI recommendation landscape is fragmented, inconsistent, and largely independent of Maps rankings.
Why AI and Maps Pick Different Businesses
The reason comes down to how each source builds its answer. Google Maps ranking is driven primarily by proximity, review count, review velocity, and business profile completeness. AI assistants, on the other hand, crawl the web, read directories, scan “best of” lists, and synthesize what they find into a recommendation.
The Awkward Media study found that the assistants lean on directories and curated lists. Perplexity cites about 19.4 sites per answer; ChatGPT about 2.7. For ChatGPT, one missing listing can decide the answer. For Perplexity, it is the weight of twenty sources.
The Nimza study found that the most-cited sites overall were winnie.com, care.com, justia.com, clutch.co, and privateschoolreview.com, depending on the trade. Google Maps and Google reviews were not among the most-cited sites for three of the four engines. Only Gemini reads Google Maps and reviews directly.
This means the directories and industry-specific platforms that AI engines cite are a different visibility channel from Maps. If you are not listed on the right directories, AI assistants may not find you at all, no matter how good your Maps ranking is.
The Lists Change, but Who Gets Left Out Does Not
A month-long study by Quaryan tracked 276 small independent businesses across six Swiss cities, asking the same 1,728 questions to ChatGPT, Gemini, Claude, and Perplexity twice, a month apart. The findings add another layer to the picture.
The lists themselves changed significantly: only about a third of the names in a city and trade were the same in both months. Perplexity’s lists changed the most. But who got named at all barely moved. 143 of the 162 businesses named in October were named again in November. 114 of 276 businesses (41%) were never named by any assistant, and 95 of those were never named in either month.
The half of a city and trade with more Google reviews was named 1.8 times as often as the other half. So review count matters for AI visibility, but it does not guarantee a mention.
The practical takeaway: a single AI answer is close to a lottery. Whether your business is in the draw at all is not random, and it changes slowly. You need to be on the right sources so AI engines can find you, and you need enough reviews and web presence that you make the shortlist.
What Local Businesses Should Do About It
Based on these three studies, here is what we recommend:
- Get listed on the directories AI engines cite. For legal services, that means Justia, Avvo, and FindLaw. For home services, Angi and Thumbtack. For healthcare, Zocdoc and Healthgrades. For child care, Winnie and Care.com. Find out which directories matter for your trade and claim your listing.
- Build your presence on “best of” lists. Many AI recommendations come from curated lists on local blogs, newspapers, and industry sites. If your city has a “best plumbers” list, you want to be on it.
- Keep collecting Google reviews. More reviews correlate with being named more often by AI assistants. The Maps ranking that comes with them is still valuable.
- Do not assume Maps ranking equals AI visibility. Track both separately. Ask ChatGPT, Gemini, and Perplexity who the best [your service] in [your city] is and see if you appear.
- Be consistent across platforms. Your business name, address, phone number, and service descriptions should match everywhere. AI engines cross-reference sources, and inconsistencies can get you dropped.
If you want help auditing your AI visibility across directories and assistants, our local SEO services include directory audit and cleanup. For a broader strategy covering both Maps and AI, check our packages or book a meeting.
Frequently Asked Questions
Do AI assistants agree with Google Maps on the best local business?
No. A study of 1,600 AI queries found that only 7% of businesses named first by AI assistants were in the Google Maps top 3, and 68% were not in the Maps top 20 at all. AI and Maps use different ranking signals and often pick completely different businesses.
Why does AI recommend different businesses than Google Maps?
Google Maps ranks businesses based on proximity, reviews, and profile completeness. AI assistants crawl the web and build recommendations from directories, best-of lists, industry sites, and other sources. Three of four major AI assistants do not read Google Maps or reviews directly.
Do AI assistant recommendations change over time?
Yes, significantly. A month-long study found that only about a third of names in a recommendation list stayed the same between months. However, which businesses get named at all changes slowly. Being in the pool of candidates is more stable than where you rank within it.
What directories should local businesses be listed on for AI visibility?
It depends on your trade. Common high-impact directories include Yelp, Angi, Thumbtack, Facebook, and industry-specific platforms like Justia for legal, Zocdoc for healthcare, and Winnie for child care. The directories AI engines cite most often vary by industry.
Does having more Google reviews help with AI recommendations?
Yes. A study of 276 businesses found that those with more Google reviews were named 1.8 times as often as those with fewer. Review count is one of several signals AI assistants use, alongside directory presence and web content.