AI search for local businesses: how ChatGPT and Gemini pick local providers

AI engines recommend local providers based on evidence they can read: your website's structure, your Google Business Profile, consistent name-address-phone data and third-party reviews. Word of mouth and a full diary are invisible to them. This guide explains how ChatGPT and Gemini answer local queries, and the specific levers that decide which businesses get named.

Ask ChatGPT to recommend an accountant, a physiotherapist or a wedding photographer in your town and it will name specific businesses. Those recommendations are not random, and they are not a copy of the Google Maps ranking. They are assembled from evidence AI engines can read. This guide covers where that evidence lives and how to strengthen it.

How do AI engines answer local queries?

Each engine assembles local answers differently, but the pattern is consistent: they retrieve web sources they trust, extract facts about businesses, and name the providers they can verify.

ChatGPT answers local queries using its search capability, which pulls from a web index and then applies its own selection criteria to what it finds. If your site does not state clearly what you do and where you operate, ChatGPT has nothing to extract. Our guide on whether ChatGPT can find your business walks through that retrieval process step by step.

Gemini is different in one important way: it has privileged access to Google’s ecosystem. It draws primarily on content Google has already indexed, and Google’s developer documentation describes Grounding with Google Maps, which connects Gemini to Google Maps data to produce location-aware answers to queries like “best Italian restaurants within a 15-minute walk”. For local businesses this means your Google presence, both your indexed website and your Business Profile, feeds directly into Gemini’s answer surface. See our guide on how to get cited in Google Gemini answers.

The common thread: AI engines do not know your reputation. They know what is written down, structured and consistent. Every lever below exists to turn your real-world reputation into machine-readable evidence.

Why is your Google Business Profile an AI search signal?

Your Google Business Profile (GBP) is the most authoritative structured record of your business that exists outside your own website. Google’s own guidance on improving local ranking names three factors: relevance (how well your profile matches the search), distance and prominence (how well known the business is, informed by links, articles and reviews).

Those same properties matter to AI answers. A complete profile gives any Maps-grounded system your category, opening hours, service area and review signal in a format it can trust. An incomplete or inconsistent one introduces doubt, and AI engines drop sources they cannot verify.

The practical work is unglamorous: claim the profile, fill in every field, keep your category precise, and follow Google’s guidelines for representing your business, which require your name, address and phone number to match the real world exactly. Businesses that treat GBP as a set-and-forget listing leave one of the few direct local AI inputs half empty.

How do reviews influence AI recommendations?

Reviews reach AI answers through three routes:

1. Prominence. Review volume and quality feed the prominence signal that decides which local businesses surface at all in Google’s ecosystem, which Gemini then draws on.
2. Third-party citations. When AI engines answer “best [service] in [town]” queries, they frequently cite roundups, directories and review platforms. If those sources rank you highly, you inherit that position in the AI answer.
3. Your own structured data. Review counts and ratings published on your site with AggregateRating markup become quotable facts. “Rated 4.9 from 212 reviews” is a machine-readable claim; “our clients love us” is not.

The SearchScore methodology scores this under Brand Authority and local platform signals: press coverage, reviews, community mentions, name-address-phone consistency and Google Business indicators are all checked in the scoring methodology.

Why does NAP consistency matter so much?

NAP (name, address, phone) consistency is how AI engines confirm that different mentions of your business refer to the same entity. If your website says “Smith & Co Accountants”, your GBP says “Smith and Co” and a directory lists an old address, an AI engine has three partial records instead of one strong one. Entity confusion is one of the quietest visibility killers, because nothing looks broken to a human visitor.

Audit every place your business is listed, directories, social profiles, professional bodies, your own footer, and make the details identical. Then declare the connections explicitly with sameAs links in your schema, as covered in our guide on brand authority signals for AI search.

What schema markup should a local business use?

Structured data is where local businesses can outrun much larger competitors, because so few implement it. The essentials:

• LocalBusiness schema (or a specific subtype such as Attorney, Dentist or AccountingService) with name, address, phone, geo coordinates, opening hours and service area.
• sameAs links pointing to your Google Business Profile, social profiles and directory listings, so engines can merge your records into one entity.
• FAQPage schema answering the questions local customers actually ask: pricing, coverage area, availability.
• AggregateRating where you legitimately display reviews.

Our guide to schema markup for AI covers implementation in detail. Across the 850,000+ websites SearchScore has audited, on-page structured data is the weakest category, averaging 23.1 out of 100 in the Q2 2026 SAVI report. Most sites have none. A local business with complete LocalBusiness markup is ahead of the vast majority of its market with a day’s work.

What does the local visibility data show?

SearchScore benchmarks entire local sectors, and the results are consistently sobering. In June 2026 we audited 151 Greater London accountancy practices across 250+ AI visibility signals for the London accountants benchmark. The average GEO score was 52.8 out of 100. Only four firms reached the Strong tier, and in only 5 of 151 cases could AI engines actually read the partner credentials that clients hire firms for.

That is the state of a competitive, professional, marketing-aware sector in a major capital city. The opportunity for a local business is not to be perfect. It is to be measurably better than a field where almost everyone is invisible.

We publish sector-by-sector guidance and live leaderboards so you can see where your industry stands:

• AI visibility for accountants and the accountancy leaderboard
• AI visibility for law firms and the legal sector leaderboard
• AI visibility for photographers
• AI visibility for therapists
• All sectors: AI visibility by industry

Where should a local business start?

Work through the levers in this order. Each one compounds the next.

1. Check where you stand. Run your site through the AI visibility checker to see engine-by-engine visibility across ChatGPT, Perplexity, Gemini, Claude, Grok and DeepSeek.
2. Fix crawler access. If AI bots are blocked in robots.txt, nothing else registers.
3. Complete your Google Business Profile. Every field, correct category, exact NAP.
4. State what you do and where, in text. Your homepage should answer “what does this business do, for whom, in which area” in its first paragraph.
5. Add LocalBusiness schema and sameAs links. Make the entity machine-readable and connect your listings.
6. Consolidate reviews and publish the numbers with legitimate rating markup.
7. Re-test and monitor. Local AI answers shift as engines update; a one-time fix decays without monitoring.

The pattern across every local sector we have benchmarked is the same: the winners are not the biggest firms, they are the ones whose websites give AI engines something concrete to cite.