How do you appear in ChatGPT's answers for local search?

There is no button and no sign-up: ChatGPT and the other AIs draw on structured public data, above all your Google Business Profile, your reviews and what the web says about you. To get cited, you need a complete, well-categorised profile, a steady flow of recent reviews, and perfect consistency of your name, address and phone number everywhere. Put simply, you do not rank on ChatGPT directly: you strengthen your Google presence, and the AI follows. The factor that separates two neighbouring businesses is still almost always the freshness and the volume of reviews.

Is GEO different from classic SEO?

It is an extension of SEO, not a separate planet. SEO tries to rank you well in a list of results that a human will scan; GEO (Generative Engine Optimization) tries to get your name into the written answer the AI offers instead. For a local business, the levers overlap almost entirely: a solid Google Business Profile, fresh reviews, replies to reviews, consistent information. Working on one strengthens the other.

Does ChatGPT really use Google reviews to recommend a business?

Generative engines do not reveal their exact recipe, and it keeps changing. But for a local recommendation, they rely on the same public data as everyone else: business profiles, mapping, reviews and the open web. Reviews play a central role because they are the richest signal about the quality and the currency of a business. A profile with plenty of recent, well-rated reviews mechanically has a better chance of being cited than a silent one.

Do you have to pay to be cited by an AI?

No, and be wary of anyone selling you a guaranteed spot in ChatGPT's answers. Unlike a sponsored slot on a marketplace, a citation in an AI answer cannot be bought: it is earned with real trust signals, fresh reviews first among them. The only investment worth making is keeping up a solid profile and a flow of honest reviews over the long run.

How many reviews do you need to be recommended by an AI?

There is no universal threshold: it all depends on your area and your direct competitors. The right instinct is not to aim for an absolute number, but to open Google Maps and ask an AI for a recommendation in your neighbourhood, then look at the businesses it cites and how many reviews they have. Your realistic goal is to get close to them, then overtake them on freshness. A business with a steady flow of recent reviews often beats a competitor that has more reviews but all of them two years old.

Does the prize wheel really help with AI search optimisation?

Indirectly but concretely, yes, because it tackles the hardest lever to sustain: the flow of fresh reviews. The customer scans a QR code in the shop, leaves their Google review in the relaxed moment of the game, then spins the wheel and wins a prize. That steady flow feeds both your Google Maps ranking and your chances of being cited by an AI. It collects no email and never makes the prize conditional on a positive review, which would be illegal: you gather honest reviews, not fake ones.

Getting cited by ChatGPT in local search: GEO for local businesses in 2026

For a few months now, a quiet shift has been changing the way people find a business near them. Instead of typing "Bastille restaurant" into Google and comparing a list, more and more people ask ChatGPT, Gemini or Perplexity directly: "can you find me a good Italian near Bastille for tonight?". And the AI does not return ten links: it cites two or three names, with a short pitch for each. So the real question for you, the business owner, becomes simple and brutal: is your place one of the names it cites, or is the AI talking about your neighbours? This article explains how the AIs choose, and what you can do right now to be in the answer.

Your customers are already asking an AI, not Google

My name is Léo, I run Pépite Pass. We equip local businesses all over France: Apple Wallet and Google Wallet loyalty cards, digital menus, and a prize wheel that helps gather Google reviews. So I spend my days buried in the Google Business Profiles of restaurants, salons and shops. And since late 2025, I have seen something new come up from the field: customers no longer search quite the way they used to.

It used to be that you typed "pizzeria near me", looked at the block of three Google profiles, and chose. Today, a growing share of people open ChatGPT or Perplexity and ask in plain language: "a good, not too pricey Italian near République for 4 people tonight?". The AI replies with two or three names, a sentence of context for each, sometimes the opening hours. It is a shortlist already sorted for you, not ten blue links to compare.

The change is enormous for a business. On Google Maps, being 4th makes you almost invisible, but at least you exist in the list. In an AI answer, there are often only two or three spots. Either you get cited, or you do not exist in the conversation at all. The game is getting harder, and almost nobody in local retail is dealing with it yet. That is exactly why there is a window to seize right now.

GEO, AI search optimisation: what exactly are we talking about

The technical name for this new territory is GEO, for Generative Engine Optimization: optimising for generative engines, meaning the AIs that write an answer instead of displaying a list of links. AI search optimisation and GEO are the same subject: getting an AI to cite your business when someone asks it for a local recommendation.

SEO and GEO, the difference in one sentence

Classic SEO tries to push your page or your profile up a list of results. GEO tries to get your name into the answer itself, the one the AI writes and the customer reads without clicking on anything. Two close logics, but not identical.

SEO: you optimise to rank well in a list that a human will scan.
GEO: you optimise to be cited and recommended inside a text that the AI writes in the human's place.
SEO: the customer sees ten options and decides. GEO: the AI has already decided, it offers two or three names.
The decisive common ground: in both cases, for a local business, everything rests on the same trust signals, with reviews on the front line.

The good news, and I will come back to this: you do not have to learn a new trade. The levers of local GEO overlap very largely with the ones you already work on (or should work on) for Google Maps. GEO is not a separate planet, it is a logical extension of your local presence.

Where AIs look to answer "a good restaurant near here"

A crucial point to grasp: when you ask an AI for a local recommendation, it does not invent the names. It has no personal opinion on the best pizza in your neighbourhood. It relies on structured data that it fetches, cross-references and summarises. And for local businesses, that data is very concentrated.

In concrete terms, the generative engines that answer a local search rely on:

Mapping and business-profile data, first and foremost Google Business Profile and Google Maps: name, category, address, hours, average rating, number of reviews.
The content and tone of recent reviews: what customers say, in which words, and on what date.
The open web: your website, the local press, directories, the blog articles that mention your place.
The consistency of all of it: if your name, your address and your activity match everywhere, the AI cites you with confidence. If it is all over the place, it ignores you to be safe.

The conclusion is direct, almost too simple: in 2026, your presence on the AIs mirrors your presence on Google. A rich, well-categorised profile with a flow of recent reviews is what feeds both your Maps ranking and your chances of being cited by an AI. You are not working on two projects, you are working on a single one that serves both.

Why fresh reviews matter as much as volume

If I had to single out one signal that decides your presence in local AI recommendations, it would be this one: the freshness of your reviews. Volume counts, of course. A profile with 200 reviews inspires more confidence than one with 12, to a human and to an algorithm alike. But volume alone is no longer enough.

Why? Because an AI recommending a restaurant for tonight wants to avoid the worst-case scenario: sending the customer to a place that has closed, changed chef, or whose quality has dropped over two years. A flow of recent reviews is the best proof that a business is alive, active, and still good right now. A profile frozen for eighteen months, even with 300 reviews, sends the opposite signal: maybe excellent, maybe closed, and the AI will not take the risk of putting it forward.

Remember the formula: volume gives credibility, freshness gives confidence. You need both. And freshness, unlike volume, cannot be bought and cannot be stockpiled: it has to be replenished constantly. That is a constraint, but above all it is an opportunity, because most of your competitors let their flow of reviews dry up after three weeks.

The concrete levers for getting cited by AIs in 2026

As promised, good news: none of this is esoteric. Here, in the order I would tackle them, are the levers that genuinely raise your chances of being cited by an AI in local search.

A complete, well-categorised Google Business Profile. It is the raw material the AI reads. A precise primary category, up-to-date hours (bank holidays included), a clear description, recent photos, filled-in attributes. Every empty field is a piece of information the AI does not have.
A flow of recent reviews, continuously. The number one signal. Not a burst once a year, a steady trickle every week. That is what proves you are alive and good right now.
Replies to reviews. Responding, to the positive and the negative alike, shows an active human behind the profile and enriches the text the AI reads about you.
Perfect consistency of your name, address and phone number everywhere on the web. One spelling, one address. Contradictory information makes the AI doubt, and it then prefers to cite a neighbour that is easier to read.
Mentions beyond your own profile: an up-to-date website, a local press article, a neighbourhood blog. The more your name appears in trustworthy contexts, the more the AI treats it as a legitimate reference.

You will notice the common thread: these are exactly the fundamentals of local Google SEO. GEO does not add a tenth project, it rewards even more strongly those who take the fundamentals seriously. If you want to dig into each building block, I have written dedicated guides:

The hardest lever to sustain, the flow of reviews: see how the wheel handles it

The review prize wheel: generating a flow of fresh reviews without cheating

The hardest lever to sustain over time is the flow of reviews. Everyone knows they should ask for reviews. Almost nobody does it regularly, because asking a customer "would you leave me a quick review?" at the till is awkward on both sides. You do it for two weeks, then you give up. And the flow dries up, at exactly the moment the AIs need it most to cite you.

That is the gap our prize wheel fills, and I want to be precise about what it does, because honesty matters to me. You display a QR code in the shop (counter, table, till). The customer scans it with their phone, with nothing to install. They go through the Google review step in the relaxed moment of the game, then spin a wheel and win a prize you have chosen. Since they have to come back to collect their prize, a scan becomes a visit, then a habit.

Two points I will not compromise on, and neither should you. One: the wheel collects no email and no phone number. It is not a contact-capture tool, it is a generator of reviews and in-store feedback, full stop. Two: the prize is never conditional on a positive review. The customer is free to leave whatever review they want, or to leave none, and they still get to play. Making a gift conditional on a good rating is banned in France (it is review manipulation, penalised by the consumer protection authorities), and in any case Google eventually detects it. You gather honest reviews, not fragile fake 5-star ones.

What GEO does not do (and the traps to avoid)

Let me be blunt, because you are going to be sold a dream about ChatGPT ranking in the coming months. Here is what local GEO is not, and the traps I would advise you not to fall into.

It is not a guaranteed spot. Nobody can promise you will be cited by ChatGPT for sure, no more than a guaranteed page 1 on Google. Run from anyone who promises it.
It is not a channel you buy. Unlike a sponsored slot on a marketplace like TheFork, you do not buy a citation in an AI answer. It has to be earned with real trust signals.
Fake reviews are a losing bet, even more than before. The AIs cross-reference sources: a surge of dubious or inconsistent reviews discredits you with the algorithm and with the model alike. Google removes them, and you lose months of reputation.
Making a gift conditional on a good rating remains banned. I repeat it because many tools do it on the sly: offering a prize only in exchange for a positive review is illegal in France and detectable. Gather reviews, not bought ratings.
The landscape moves fast. The AIs change their sources and their methods every few months. The only robust strategy is to bet on durable fundamentals (a solid profile, fresh and honest reviews) rather than on a trick of the moment.

The rule of conduct is simple: play the long game. The business owner who builds a solid profile and a flow of honest reviews will not only rank well on Maps, they will become the safe bet the AIs cite because there is no risk in recommending them.

Where to start this week

Enough theory. If I took over a business tomorrow with the aim of being cited by the AIs as much as by Google Maps, here is the exact order I would go about it.

Day 1: the profile. A precise primary category, every field filled in, a dozen recent photos, exact hours. It is the raw material the AI reads about you.
Day 2: consistency. Settle on one official spelling of your name, one address, one number, and clean up the old profiles and directories that diverge.
Day 3: observation. Ask ChatGPT and Google Maps the question the way a customer in your neighbourhood would. Note who gets cited, with how many reviews and how fresh they are. That is your real target.
From day 4 onwards: the flow of reviews. Put in place a regular mechanism to gather fresh reviews every week, indefinitely. It is the only genuinely durable lever, and the one almost all your competitors lack.
Continuously: reply to every review in a few words. A free activity signal, read by Google and by the AIs.

The common thread in all of this is that a flow of fresh, honest reviews is the shared fuel for your visibility on Google Maps and for your citations by the AIs. That is precisely what Pépite Pass automates with the prize wheel, with no app to install, no email collected, and never any prize tied to a good rating. You can try it for free, no credit card, and judge on your own reviews. The window to get ahead on this is open now, before all your neighbours get to it.