South Florida · Mexico City The OGs of digital marketing · now fluent in AI ● AI systems in production
Journal · AI Marketing

Restaurants in Mexico: How Diners Find You in Google and AI Answers

A Mexican restaurant is easy for Google and ChatGPT to find but often gives them little to say. In our October 2026 check of 15 Mexico City restaurant and café homepages, none blocked ChatGPT’s search crawler, yet all 15 lacked content that answers a diner’s questions and 12 had a missing or badly sized meta description.

Here is how diners search, what the numbers show, and the order we would fix things in.

How crowded is the restaurant market in Mexico?

Very. Data México, the Ministry of Economy’s open-data site, reports 740,231 economic units in food and drink preparation, drawn from INEGI’s DENUE directory for May 2026. Estado de México has 89,408, Ciudad de México 55,753 and Jalisco 48,957. That count includes everything from a taco stand to a tasting-menu room, so it overstates the number of businesses you actually compete with. It still shows why “a good restaurant” is not a position anyone can claim by being open.

Diners reach those businesses through a phone. INEGI’s 2025 household technology survey, published 16 June 2026, found that 86.1% of people aged 6 and over used the internet and 84.6% used a mobile phone. In Mexico City, 90.5% of households had internet. Of smartphone users, 90.6% used instant-messaging apps.

The sources do not say how many of those users ask an AI assistant where to eat. We will not guess. The practical point is simpler: whichever screen the diner uses, the facts it shows come from places you control or can influence.

Where does a hungry person start looking?

Google’s help page says that when someone searches near their location, they find local results on Google Maps and Search, and its own example is “Italian restaurant.” The page says results are mainly based on relevance, distance and popularity, and that there is “no way to request or pay for a better local ranking.” Google describes the third factor as prominence, meaning how well-known a business is, and bases it partly on how many websites link to you and how many reviews you have.

The same page lists what you can control: verify the profile, keep it accurate and complete (address, hours including special hours, category, attributes such as parking or Wi-Fi), reply to reviews and add photos and video.

Google’s guide to generative AI features (last updated 10 July 2026) connects this to AI answers. It says generative responses can include information about local businesses, and that Business Profiles help those businesses be visible in AI responses and other Search results. Our reading: the AI answer is not a separate game. It leans on the same listing facts, so a stale Business Profile is wrong in both places.

What did 15 Mexico City restaurant homepages show?

In October 2026 we ran our 23-check homepage test on 112 businesses in Mexico City, including 15 restaurants and cafés. We report totals only and name no one. The restaurant group averaged 16.1 of 23 checks passed, with scores from 11 to 21.

Check (homepage, plain request)Restaurants failing (of 15)
Blocks ChatGPT’s search crawler0
No content that answers customer questions15
Meta description missing or wrong length12
Too little readable text10
Not exactly one H19
No contact form on the homepage11
No tap-to-call or WhatsApp link5
No business (LocalBusiness-type) markup5
No structured data at all5

Across all 112 businesses, 103 (92%) had no content answering customer questions, so restaurants are not alone. They are simply the group where the gap was universal in our sample.

Why is the menu the weak point?

Ten of the 15 homepages had too little readable text for our check to work with. That does not prove the menu is missing. Our reading is that on many restaurant sites the menu, the story and the prices live in an image, a PDF or a script, which a plain request cannot read and a crawler may not either.

A diner (or an assistant answering for one) asks things like: is there a vegetarian option, how much is the tasting menu, do you take groups of ten, is there parking. If the answer sits in a photo of a paper menu, the assistant has nothing to quote and falls back to whatever a review or a delivery listing says, which may be out of date.

Schema.org gives restaurants specific fields for this: hasMenu (the menu as structured data, text or a URL), servesCuisine and acceptsReservations. Google’s local business documentation, updated 8 September 2026, also shows opening hours and price range in its restaurant example. It notes that the Restaurant carousel is limited to a small set of providers, so do not build a plan around getting one.

What should a restaurant fix first?

This is the order we would work in. Step one costs nothing and fixes the most visible listing.

  1. Verify and complete the Business Profile. Category, address, regular and special hours, attributes, photos. Reply to reviews.
  2. Put the address, hours and phone in plain text on the homepage, not only in a banner image.
  3. Publish the menu as a web page, with dish names and prices as text. Keep the PDF if you like, but as a second option.
  4. Give every location its own page with its own address, hours and map.
  5. Add a reservation or WhatsApp route a thumb can reach: tap-to-call, a WhatsApp link and, if you take bookings, a short form.
  6. Add Restaurant markup with name, address, phone, cuisine, price range, hours, menu link and reservation link, emitted as one clean block.
  7. Write ten real answers under question headings: parking, groups, dietary needs, dress code, delivery, private events, opening on holidays. Use the questions your staff actually hear.

On the last step, Google’s generative AI guide asks for what it calls non-commodity content: a unique point of view, written from first-hand knowledge rather than a summary of what is already online. A restaurant has that material by default: how a dish is made, where the produce comes from, which hours are quiet and which are busy. Say it plainly, using the dish names and the neighbourhood a diner would type, and keep it current when the menu changes. A page that was right in 2024 and wrong today is worse than no page, because an assistant may repeat it.

If you take reservations by message, an automated follow-up through GoHighLevel keeps an answer going out within minutes even when the kitchen is busy.

Should a restaurant let ChatGPT’s crawler in?

Yes, unless you have a reason not to. OpenAI’s crawler documentation says OAI-SearchBot is used to surface websites in ChatGPT’s search features, and that sites opted out of it “will not be shown in ChatGPT search answers,” though they can still appear as navigational links. It is a separate setting from GPTBot, which relates to training, so you can allow one and not the other.

Allowing it does not guarantee a mention; it only removes one way of being left out. None of the 15 restaurants in our sample blocked it, so for most owners this is already done. Check your robots.txt anyway: a new website plugin can change it without anyone noticing. Our guide to showing up in ChatGPT covers the other crawlers, and the free AI Visibility Checker tests yours in a minute.

What does this check not tell us?

It is a snapshot of 15 Mexico City homepages from two sets, read once, without running scripts. The sample is small and not random, and it covers restaurants that have a website, so it says nothing about the many that live only on Instagram or a delivery app. We did not test whether any assistant recommends these restaurants, and nobody should sell you that guarantee. Google itself says that optimizing for its generative AI features is still SEO.

What we can say is narrower and useful: the crawlers are not the obstacle. The missing piece is plain, current, quotable information. That is the focus of our restaurant marketing work (in Spanish: marketing para restaurantes), and the same thinking sits behind our generative engine optimization guide.

Want your business found on Google and in ChatGPT? Ciao Bella engagements start at $5,000 a month, no setup fee. See how we work.

Verified 26 October 2026 against INEGI, Data México (Secretaría de Economía), Google Search Central, Google Business Profile Help, OpenAI’s crawler documentation, Schema.org and our own check of 15 CDMX restaurant homepages

Building on the web since 1999 · South Florida and Mexico City

Frequently asked questions

Do restaurants need a website if they have Google Maps and Instagram?

Not to appear on the map, but a website you control gives Google and AI assistants one current source for your menu, hours and address. In our sample, 15 of 15 homepages still had nothing answering diner questions, so a basic, accurate site is an open advantage.

Is a PDF menu bad for Google and ChatGPT?

A PDF can be read, but it is slower to use on a phone and easy to leave out of date. Publish the menu as a web page with text and prices, and keep the PDF as an extra.

What structured data should a restaurant use?

Restaurant markup with name, address, phone, cuisine, price range, opening hours, a menu link and a reservation link. Google says structured data is not required for its AI features, but it helps with rich results.

Can I pay to rank higher in Google Maps?

Google states there is no way to request or pay for a better local ranking. Relevance, distance and popularity decide, and you can improve the first and third by keeping your profile complete and earning genuine reviews.

How can I check my own restaurant website?

Run its address through the free AI Visibility Checker on this site. It uses the same 23 checks as this study and lists what to fix.

Sources

  1. Data México (Secretaría de Economía) — Servicios de preparación de alimentos y bebidas DENUE May 2026: 740,231 economic units; Estado de México 89,408, Ciudad de México 55,753, Jalisco 48,957.
  2. INEGI — ENDUTIH 2025, press release 32/26 (16 June 2026) 86.1% of people aged 6+ used the internet; 84.6% used a mobile phone; 90.5% of households in Mexico City had internet.
  3. INEGI — ENDUTIH 2025, results report 19/26 (16 June 2026) 90.6% of smartphone users used instant-messaging apps.
  4. Google Business Profile Help — Tips to improve your local ranking on Google Local results are mainly based on relevance, distance and popularity; no way to request or pay for a better ranking.
  5. Google Search Central — Optimizing your website for generative AI features on Google Search Last updated 10 July 2026. Business Profiles help local businesses be visible in AI responses; optimizing for generative AI search is still SEO.
  6. Google Search Central — Local business structured data Last updated 8 September 2026. Restaurant example with menu, cuisine, hours and price range; the Restaurant carousel is limited to a small set of providers.
  7. OpenAI — Overview of OpenAI Crawlers OAI-SearchBot surfaces sites in ChatGPT search; opted-out sites will not be shown in ChatGPT search answers; independent of GPTBot.
  8. Schema.org — Restaurant Properties hasMenu, servesCuisine, acceptsReservations.

Related reading: for the wider picture of how we help restaurants, see restaurant marketing and, in Spanish, marketing para restaurantes. For how AI search works in practice, read generative engine optimization, and to see how your own homepage scores, use the free AI Visibility Checker.

Want diners to find your restaurant by name?

Tell us your site — we’ll run the same checks and send what we find.

Want to see what it costs first? Engagements start at $5,000 a month. See how we work.

We use your details only to reply and talk about working together. We never sell them. Privacy policy.

Free newsletter

Get our AI-search notes by email.

One short take a day: what changed in Google and ChatGPT, and what to do about it. Unsubscribe anytime.

No spam. We've been online since 1999. We know what spam looks like.

Hey, want to chat?
Scroll to Top