Local GEO: how to get named in AI answers
Summary
Ask four AI engines who the best local business is, and they usually name four different ones. Here's how to get picked.
Ask four AI engines the same question. “What are the top 10 dentists near Park Slope?” You will usually get four different winners. Not a reshuffle of the same list. Different businesses.
Yext asked Gemini, OpenAI, Perplexity, and Anthropic that kind of question almost half a million times. All four put the same business first in 4% of cases. In 48%, each engine picked someone else. I found the write-up through Alex Groberman.
If you own a local business, that sounds grim. I read it the other way. Almost nobody owns the answer yet. The businesses that show up on all four have a few things in common, and you can copy them.
A Maps ranking is not an AI answer
Local GEO, sometimes called local AEO, is getting your business named in an AI answer. ChatGPT, Perplexity, Gemini, Google’s AI Overviews and AI Mode, Claude, Copilot, Grok.
It overlaps with local SEO. It is not the same job. Local SEO is a list of links and map pins. An AI answer is a shortlist the model assembles after running several searches behind the scenes. You can do well in Maps and still be missing from the answer.
A few figures are worth keeping. They all point at the same job.
- Each engine named about 7 businesses, even when asked for 10. One question surfaced 23 different businesses on average.
- They don’t cite the same pages. Gemini mostly cited the business’s own site (42%). OpenAI leaned on directories (41%). Anthropic used review sites far more than the others (21%).
- When all four did agree, the winner’s own site was cited 91% of the time, and those businesses had a median of 568 reviews, against 19. The highest-rated name still came first only 40% of the time. That 91% is only those rare agreements. A website does not win you 91% of questions.
- BrightLocal looked at more than 200,000 local AI searches. Business websites were 42% of citations. The tracked business showed up in Maps on 66% of searches, and in the AI answer on 32–38%. AI Mode and AI Overviews shared about 29% of the businesses they named. ChatGPT overlapped with either only about 20%. Google Business Profile was the biggest source, and Yelp showed up in 80% of ChatGPT’s local answers.
If an engine never puts you in the pile, your rating never gets a vote.
Each engine reads a different shelf
They are not reading the same shelf.
- Google’s AI Overviews and AI Mode lean on Google Business Profile. That’s from BrightLocal’s source breakdown.
- ChatGPT sends the search out to partners. OpenAI’s help center says it rewrites the query, and it uses your rough location for “near me” questions. In July 2026, Yelp licensed its reviews, photos, and business data to OpenAI.
- Copilot sits on Bing. Bing Webmaster Tools has an AI Performance report that sorts the queries behind your citations, including a Local intent.
- Apple (Siri, Maps, Spotlight) uses its own place data. In 2026 it folded Business Connect into Apple Business, which is where you claim how your location appears.
There isn’t one trick that covers all of them. Be complete, and be the same business, everywhere they look.
Fill in the profile each engine reads
- Google Business Profile. The right primary category, services, hours, attributes, a description, photos, and the occasional post. Category and accurate hours are near the top of Whitespark’s Maps factors.
- Bing Places, because Copilot reads Bing.
- Apple Business, for Siri, Apple Maps, and Spotlight.
- Yelp, because of the OpenAI deal, plus the directories people in your category actually use. Tripadvisor, OpenTable, Angi, BBB, and the healthcare or legal ones if that’s your work.
- Keep the name, address, and phone number identical. Joy Hawkins has pointed out that ChatGPT still gets local details wrong, and it pulls from odd places, including MapQuest. A messy listing gets repeated.
Pages a model can quote
Make the website obvious
Both studies land here first. A model has to answer plain questions from your pages. Where you are. What you do. Which areas you serve. Roughly what it costs. Why someone should trust you.
- One page per service, and one page per location. In Whitespark’s 2026 Local Search Ranking Factors, a dedicated service page was the second-strongest factor for AI search visibility.
- Answer the questions customers actually type. “Do you take emergency calls on weekends?” “Do you take my insurance?”
- Put your name, address, and phone number in the HTML, matching your Google Business Profile.
Add LocalBusiness schema
On each location page, add LocalBusiness schema: name, address, phone, hours, coordinates, and sameAs links to your real profiles. In the Yext data, schema lined up with a 2.1× higher chance of being named. It will not get you recommended on its own. It makes the facts easier to read.
Reviews that keep arriving
Count and a steady pace mattered more than the star rating.
The businesses all four engines agreed on were the ones with hundreds of reviews, about a dozen new ones a month, and replies on most of them. Insites looked at 10,000 US local businesses and found a similar gap: businesses named by both ChatGPT and Perplexity averaged 133 Google reviews. Businesses named by neither averaged 11.
Reviews matter more once someone is close to buying. Seer studied about 800,000 AI responses and found review sites’ share of citations rose from about 1.5% on early questions to about 24% when the question showed intent to buy. A thin third-party profile reached a 53.5% median citation rate, against 1% with none. ChatGPT will also say when you have no third-party reviews, and repeat the complaints it finds.
Ask steadily, not in one burst. Collect them somewhere besides Google. Reply, including to the bad ones. Then read them the way a model will summarize them, because that summary is what a lot of people will see.
Someone else has to say it
In the Whitespark survey, the strongest AI visibility factor was a place on expert “best of” lists. Three of the top five factors were about other people talking about you: industry sites, local news and blogs and associations, and the quality of the review sites you’re on.
- Ask to be in real “best plumber in Hoboken” roundups. Skip the ones you can buy.
- Local press, the chamber, a community event, a blogger who already covers your city. Chamber pages showed up 1.8× more often in Yext’s unanimous answers than in the split ones.
- Turn up on Reddit, YouTube, and forums as someone who knows the work. Joy Hawkins makes the same point as more of the web becomes generated text.
Check more than ChatGPT
One look at ChatGPT tells you almost nothing. The answer changes between runs, between engines, and between neighborhoods in the same city. BrightLocal found ChatGPT’s results overlapped only about 36% when the same prompt was run from different points in one town.
- Write 10 to 30 prompts the way customers ask. “Emergency plumber open now in Astoria.” “Best kid-friendly brunch in Hoboken.” Put the place in the prompt.
- Run them on the engines your customers use, on a schedule.
- Note whether you were named, where you sat, and which pages got cited. Then look at who keeps showing up instead of you.
What to have in place
- A page for each service and each location, with your name, address, and phone number in the HTML
- The questions customers actually ask, answered on those pages
- LocalBusiness schema on every location page, with links to your real profiles
- Google Business Profile filled in: category, services, hours, attributes, description, photos, a recent post
- Bing Places claimed, and matching
- Apple Business claimed, for Maps and Siri
- Yelp and the directories in your category, claimed, with the same details
- The same name, address, and phone number everywhere, including the data aggregators
- New reviews coming in on Google and at least one other site
- Replies on most reviews, including the negative ones
- A few real “best of” lists or local press mentions
- AI crawlers allowed through, including OAI-SearchBot, in robots.txt and at the CDN
- A fixed set of local prompts, checked across the engines your customers use
Common questions
Is this different from local SEO?
They overlap, and the results don’t match. BrightLocal saw the tracked business in Maps on 66% of searches, and in the AI answers on 32–38%. A strong Maps ranking helps. It does not get you named.
Why does ChatGPT pick someone Google doesn’t?
They read different sources. Google’s AI features lean on the Business Profile. ChatGPT pulls from outside search partners and from directories like Yelp. In the Yext data, about 41% of OpenAI’s citations were directories. About 42% of Gemini’s were the business’s own site.
Do reviews change who gets recommended?
Yes, in every dataset I looked at. How many you have, and how recently they arrived, mattered more than the star rating. The highest-rated name in a list came first only 40% of the time.
Does schema help?
A bit. Yext tied it to a 2.1× higher chance of being named. Treat it as a way to make your details easy to read, not as a trick.
How do I check?
Write the prompts in your customers’ words, include the neighborhood, and run them on several engines for a few weeks. One session will mislead you. The answers move.
When four engines agree on a winner 4% of the time, nobody owns the answer. That includes your competitors. The businesses that do show up everywhere are not doing anything exotic. A clear website, finished profiles, reviews that keep coming, and a few other sites willing to vouch for them.
The step people skip is checking. We built ChatBakers to watch how a brand shows up across ChatGPT, Perplexity, Gemini, Google AI Overviews, Claude, Copilot, and Grok: which prompts name you, where you sit, and which sources the answer cited. If you’re working on this locally, pick a set of city prompts before you change anything, and watch what each engine does. Measure it.
Sources
- Yext Research, Four AIs, one question, four different answers (Feb–Aug 2026 data); fact pages: 4% agreement, 91% own-site citation
- Alex Groberman on X: post 1, post 2
- BrightLocal, What 200k local AI searches tell us (Sep 16, 2026) and Where to get local citations for AI search
- Whitespark, 2026 Local Search Ranking Factors
- Seer Interactive, Study of 800K AI Responses: How Review Profiles Shape Brand Presence in AI Search (May 2026, commissioned by Trustpilot)
- Insites, Do Google reviews affect AI recommendations? (Aug 2026)
- Sterling Sky (Joy Hawkins), The State of Local SEO in 2026; Uberall Local Marketing Beat, Episode 29 with Joy Hawkins
- Microsoft Bing, New AI Visibility Insights in Bing Webmaster Tools (Jun 16, 2026)
- OpenAI Help Center, Searching the web with ChatGPT
- Axios, Yelp partners with OpenAI to surface reviews in ChatGPT (Jul 23, 2026)
- Apple Newsroom, Introducing Apple Business (Mar 24, 2026)
- Google Search Central, LocalBusiness structured data
Monitor your AI visibility
See how AI models talk about your brand. Start your free 7-day trial.
Start Free Trial →