How to Show Up in ChatGPT and AI Search as a Local Business

To show up in AI answers as a local business, you optimize the sources the engines actually read: your own website, business listings and review profiles. Across 6.8 million AI citations analyzed by Yext, 86% came from sources the brand already controls. Each engine has its own diet — ChatGPT leans on Bing and directories like Yelp and Foursquare, Gemini grounds its answers in Google Maps, Perplexity integrates Yelp data and has a Tripadvisor partnership — so the real work is consistency across all of them, not one secret trick.

How many people already ask AI for local recommendations?

45% of consumers used generative AI tools such as ChatGPT to get local business recommendations in the past year, up from 6% in 2025, according to BrightLocal’s Local Consumer Review Survey 2026. That makes AI the third source of local recommendations, behind only Google (71%) and Facebook. And 82% of consumers now read AI-generated summaries of reviews.

Google’s own results are changing too. A Local Falcon whitepaper (May 2025; 60,000 queries, 4,423 businesses, 20 countries) measured that 40.2% of local business queries triggered an AI Overview in April 2025. The pattern matches the broader shift we track in AI search adoption: the question is no longer whether customers ask AI about businesses like yours — it is what the AI answers.

Where does each AI engine get its local answers?

Each engine grounds local answers in a different data supply, so a shop or hotel can be visible in one and absent from another. There is no single “AI ranking” — there are at least three, each with its own sources.

Engine	Local grounding	Key local sources	Contact data accuracy (SOCi, 2026)
ChatGPT	Bing index / Bing Places	Directories: 48.7% of its citations (Yext); Foursquare, Yelp; Facebook for reviews	~68%
Perplexity	Data partnerships and integrations	Yelp (integrated 2024), Tripadvisor (partnership 2025), MapQuest	~68%
Gemini	Google Maps grounding, 250M+ places	Business websites: 52.1% of its citations (Yext); Google reviews	100%

The details behind the table. Whitespark (September 2025; 153 local queries, 17 categories, 9 US cities) found ChatGPT’s local recommendations rest on Bing’s index. Perplexity integrated Yelp’s maps, reviews and business listings in March 2024, then partnered with Tripadvisor in January 2025: one billion reviews and contributions, more than 11 million business listings across 43 markets and 22 languages. Google launched Grounding with Google Maps in the Gemini API in October 2025, anchoring answers in data from more than 250 million places, including reviews, hours and photos.

On ChatGPT’s supply chain, OpenAI’s help documentation states — as reported in secondary sources quoting it, since the page did not load for us directly — that ChatGPT search may share queries, disassociated from the account, with Bing and other third-party search and data providers; Bing and Shopify are the only providers it names.

Which sources do AI engines actually cite for local businesses?

Mostly sources the business already controls. The largest dataset available is Yext’s study (October 2025): 6.8 million AI citations across ChatGPT, Gemini and Perplexity, from 1.6 million queries per model.

“86% of AI Citations Come from Brand-Managed Sources” — Yext, October 2025

The breakdown: brands’ own websites 44% (2.9 million citations), listings and directories 42% (2.9 million), reviews and social 8% (545,000). Reddit and forums, despite the hype, only around 2%.

The mix also moves fast. BrightLocal’s first ChatGPT study (December 2024; 800 manual searches across 20 sectors and 20 US cities) found 58% business websites, 27% mentions and 15% directories — with Reddit and Quora appearing zero times. By July 2025, a follow-up BrightLocal study — repeated searches across Google AI Mode, Gemini, Perplexity and ChatGPT Search in 10 sectors — found Yelp cited in 33% of all searches, with Perplexity using it in every sector, and 60-70% of ChatGPT’s local results deriving from Foursquare listings, especially in smaller cities and niche categories — a smaller sample, so treat the exact figure with caution. Same vendor, seven months apart: in December 2024 Yelp did not appear as a directory source across 800 ChatGPT searches; by July 2025 it was a source in a third of a four-engine search set. The two are not like-for-like measurements, but they point the same way — source diets change fast, which is why one-off audits age quickly. The mechanics behind these choices are covered in how AI engines pick their citations.

Does Google Business Profile still matter if ChatGPT ignores it?

Yes — but for Google’s AI, not for ChatGPT. BrightLocal’s December 2024 study is blunt about what ChatGPT did not use:

“Yelp, Facebook, and Google Maps do not appear as directory sources within this study” — BrightLocal, December 2024

Whitespark’s September 2025 research points the same way: ChatGPT’s local answers sit on Bing Places, and Google Business Profile reviews were absent from the dominant positions. Meanwhile Google explicitly recommends keeping the Business Profile up to date for its AI features, and Gemini grounds its local answers in Google Maps data covering more than 250 million places — which is why SOCi’s index measured Gemini’s contact data at 100% accuracy.

The practical translation: keep Google Business Profile current — it is the data layer Gemini reads and the one Google itself points local businesses to. Bing Places is the equivalent homework for ChatGPT, and in our experience it is the one almost no small business has claimed (ASSUMED, from our own audits — no study measures it).

How much do reviews weigh in AI recommendations?

Reviews act as a quality filter — and not only Google’s reviews. In SOCi’s Local Visibility Index 2026 (reported by Search Engine Land, January 2026), locations recommended by ChatGPT average 4.3 stars; Perplexity’s picks average 4.1 and Gemini’s 3.9.

Where the reviews live matters per engine. Whitespark found Facebook was ChatGPT’s most frequent review source, followed by Yelp — cited more than twice as often as the third place — and TripAdvisor. Perplexity surfaces Yelp and Tripadvisor reviews through its data deals. Gemini reads Google Maps reviews. Add BrightLocal’s finding that 82% of consumers read AI-generated review summaries, and the conclusion writes itself: machines are already summarizing your reputation, whether you manage it or not.

Does LocalBusiness schema get you into AI answers?

Not by itself — and Google says so explicitly. Its guide to appearing in AI features (updated July 2026) states that structured data is not a requirement for AI features and that Google Search ignores llms.txt files:

“You don’t need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search” — Google Search Central, July 2026

What the LocalBusiness documentation does say: only name and address are required, the most specific subtype (Restaurant, DaySpa, Electrician) is recommended, and the markup enables the knowledge panel. No AI benefit is promised anywhere in the document. The honest case for schema is consistency: machine-readable name, address and opening hours that every engine can verify against your listings. It is one small piece of generative engine optimization, not a shortcut.

Is AI visibility harder than ranking in Google’s local pack?

Much harder, by current measurement. SOCi’s Local Visibility Index 2026 covered almost 350,000 locations from 2,751 multi-location brands: ChatGPT recommended only 1.2% of locations, against the 35.9% that appear in Google’s local 3-pack. Gemini recommended 11% and Perplexity 7.4%. Search Engine Land sums it up as AI visibility being three to 30 times harder to achieve than traditional local ranking.

Two honest caveats. SOCi measures multi-location brands, so extrapolate to independent small businesses with care. And there is a genuine opening: Local Falcon found effectively no correlation between distance to the business and position inside an AI Overview (coefficient 0.001). Unlike the local pack, proximity barely counts — a well-documented small business can be recommended beyond its traditional radius, and an established competitor gets no geographic shield.

Does this US data apply to a local business in Europe?

The mechanisms travel; the directory mix probably does not. Every citation study above ran on US cities and US brands — no equivalent study has been published for Italy, Spain or Slovenia. Bing’s index, Perplexity’s Yelp and Tripadvisor deals, and Gemini’s Maps grounding are global mechanisms: those are measured. The exact weight of each directory in a European market is an assumption, not a measurement — Yelp’s footprint is thin in much of Europe, where TripAdvisor, national directories and Google itself likely carry more weight.

In our labeling system, that distinction matters: the mechanisms are MEASURED, the European directory mix is ASSUMED. Any vendor selling you certainty about it deserves suspicion. The only way to know is to test prompts in your own language and market.

What should your local business do this week?

Everything above reduces to a short, mostly free checklist:

1. Fix your NAP. Name, address, phone and hours identical on your website, in plain HTML text, on every page that mentions them.
2. Claim both profile systems. Google Business Profile feeds Gemini’s Maps grounding and is Google’s own advice for AI features; Bing Places feeds ChatGPT. Both are free.
3. Cover the directories AI actually reads. Foursquare, Yelp where relevant, TripAdvisor for hospitality, Facebook — the sources measured by BrightLocal and Whitespark.
4. Add LocalBusiness schema with the most specific subtype. Not for an AI boost Google never promised, but for verifiable consistency.
5. Work reviews beyond Google. ChatGPT’s picks average 4.3 stars, and it reads Facebook, Yelp and TripAdvisor first.
6. Measure before and after. Ask ChatGPT, Perplexity and Gemini 10-15 real customer questions. Record whether you appear and whether your data is right — contact details are only ~68% accurate on ChatGPT and Perplexity, per SOCi. Re-test quarterly, because source diets shift.

That baseline-first sequence — measure, fix, re-measure — is exactly how we run an AI visibility audit for small businesses: no invented metrics, just what the engines say about you before and after.