How to Get Your Restaurant Recommended by ChatGPT (Step by Step)

To get your restaurant recommended by ChatGPT, let OAI-SearchBot crawl your site, publish your menu as text with prices, add Restaurant structured data, complete your Google Business Profile, and earn reviews where ChatGPT looks.

This is the how-to: what ChatGPT, Perplexity, and Google's AI features say they read, in their own documentation, then seven steps in order, what to skip, and a structured data block you can paste. Measuring whether it worked is covered separately in how to track your restaurant's AI search visibility.

Where does ChatGPT get its restaurant information?

ChatGPT gets restaurant information from live web search: OpenAI says ChatGPT search uses third-party search providers plus content from its partners, crawls sites with OAI-SearchBot, uses your approximate location for local results, and shows reservation availability from OpenTable, Resy, or Yelp.

OpenAI's help article Searching the web with ChatGPT is specific about restaurants. ChatGPT "can also use location information to find local results, including restaurants, news, and weather." When it partners with search providers, it rewrites the question into targeted queries: the page's own example turns "What are some good restaurants near me?" from a San Francisco user into "top restaurants San Francisco." For how those providers handle queries, the page points to Microsoft's and Shopify's privacy policies, which is as close as OpenAI comes to naming them.

Three more lines from the same page:

• Listings come from partners. ChatGPT "may share your location with trusted third-party providers that help provide more local information, such as nearby business listings." On iOS and Android, results may include a map.
• Reservations come from OpenTable, Resy, or Yelp, "depending on the restaurant." OpenTable is available globally, Resy in the United States, Yelp in the United States and Canada.
• Eligibility is simple, ranking is not. "ChatGPT ranks search results using multiple factors intended to help users find relevant, reliable information. Placement is not guaranteed." To be eligible, "allow OAI-Searchbot to crawl the site and confirm that the website host or content delivery network allows traffic from OpenAI's published searchbot IP addresses."

OpenAI's crawler documentation splits the bots by job: OAI-SearchBot is "used to surface websites in search results in ChatGPT's search features" and is not used for training, GPTBot crawls for model training, and ChatGPT-User fetches a page when a user asks for it. A robots.txt change takes about 24 hours to register. In plain terms, ChatGPT runs searches the way a person would, reads what it finds, and pulls listings and booking data from partners.

How do Perplexity and Google AI Overviews choose restaurants?

Perplexity surfaces and links websites crawled by PerplexityBot, and Google's AI Overviews and AI Mode draw on Google's normal search index plus your Business Profile, with no requirements beyond being indexed and snippet-eligible, according to each company's own documentation.

Perplexity. Its crawler docs say PerplexityBot "is designed to surface and link websites in search results on Perplexity" and "is not used to crawl content for AI foundation models." A second agent, Perplexity-User, visits a page when a user's question needs it and "generally ignores robots.txt rules." No submission process exists: if PerplexityBot can crawl the page, it can be cited.

Google AI Overviews and AI Mode. Google's page on AI features and your website says: "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." To be cited, "a page must be indexed and eligible to be shown in Google Search with a snippet." Its two restaurant-relevant tips: keep Business Profile (and Merchant Center) information up to date, and make sure structured data matches the visible text on the page.

One trap: Google-Extended. Per Google's crawler list, it controls Gemini training and grounding only and "does not impact a site's inclusion in Google Search nor is it used as a ranking signal." Blocking it does not remove you from AI Overviews. Blocking Googlebot does.

Which sources do AI assistants cite for "best tacos near me"?

No assistant publishes its source mix, but their documentation points to the same places: your website, Google Business Profile data, the reservation and review platforms they name (OpenTable, Resy, Yelp), and whatever ranks in ordinary web search, which for local food means review sites, local press, and forum threads.

The weighting is not published and shifts with phrasing, location, and (for ChatGPT) what the user has told it about themselves. OpenAI warns that "search results and citations can be incomplete, outdated, or incorrect." So make every source say the same thing about you, then test your own questions; the method is in the tracking post.

How do I get my restaurant on ChatGPT, step by step?

Seven steps, in order: open your site to the crawlers, publish the menu as text, add Restaurant structured data, complete your Google Business Profile, get listed and reviewed on the partner sites, answer diner questions on your own pages, and test ten queries a month.

Step 1: Let the crawlers in

robots.txt sits at the root of your domain. Google's robots.txt guide gives the two rules you need: "the default assumption is that a user agent can crawl any page or directory not blocked by a disallow rule," and a crawler follows only "the first, most specific group that matches" its user agent. So a missing file, or one that only says User-agent: * and Allow: /, blocks nobody. Problems come from a Disallow: / left over from launch or a "block AI bots" toggle in a security plugin or CDN. Open yourrestaurant.com/robots.txt and look for these names.

Then the host side: OpenAI asks you to confirm your host or CDN allows its published searchbot IP addresses, and a "block AI bots" toggle usually blocks the search crawlers along with the training ones.

Step 2: Put the menu on your site as text, with prices

Every item as real text: name, one-line description, price, and any flag a diner asks about (vegetarian, gluten-free, contains nuts). One page per menu with a stable URL. Not a PDF, not a photo. Google's Business Profile rules say a menu should list "the complete set of food and drink items that are available at the business," and that "menu URLs can't be direct links to third-party ordering or delivery services," so the menu page belongs on your own domain, not on a marketplace. If your site cannot do that today, the direct ordering website guide covers what a rebuild needs.

Step 3: Add Restaurant structured data

Structured data is a small JSON-LD block in the page's HTML that states your facts in machine-readable form. Google's local business documentation requires name and address, and recommends geo coordinates (at least 5 decimal places), openingHoursSpecification, priceRange, servesCuisine, telephone, url, and menu, "the fully-qualified URL of the menu." The schema.org Restaurant type adds hasMenu and acceptsReservations. Google's rule applies: the markup must match what a visitor sees. A copy-ready example is below.

Step 4: Keep your Google Business Profile complete and consistent

The Business Profile guidelines are strict. Use the name your restaurant is known by in the real world; Google says adding unnecessary information to the name "isn't permitted, and could result in the suspension of your Business Profile." Use a real location and a local phone number. For hours, Google's restaurant rule: "Use the hours when diners can sit down and dine in your restaurant. Otherwise, use takeout hours." Pick categories that complete "This business IS a," not "HAS a." Add the menu link (on your domain), current photos, and your own ordering link set as preferred; Google's food ordering link rules say providers "must remove third-party links from a profile within 5 business days" once you ask, and the click path is in how to add an order button on Google. Then make name, address, and phone match your website and every other listing exactly.

Step 5: Get listed and reviewed where the assistants look

Claim the listings the assistants name. Yelp says it is free to be on Yelp; a claimed page lets you update business info, add photos, and respond to reviews and messages. Tripadvisor offers a free business listing claim with the same controls. If you take reservations, be on at least one provider ChatGPT shows availability from (OpenTable, Resy, or Yelp).

Reviews follow each platform's rules. Google's policy bans incentives and asking only happy customers (next section). Yelp's content guidelines say "Businesses should never ask customers to write reviews." Everywhere, reply to reviews, especially bad ones, in plain language.

Step 6: Write pages that answer the questions people ask assistants

People ask assistants what they used to phone in: "Do you have gluten-free options?" "Is there parking?" "Do you deliver to Maple Heights?" Put a page on your site that answers each one in its first sentence, in plain words, and keep it true. One FAQ or Visit page with 10 to 15 questions is enough. Make the answers match your structured data and Business Profile (same hours, same delivery zone, same phone). Ranking locally beyond that is covered in local SEO for restaurants.

Step 7: Test ten questions a month

Once a month, ask ChatGPT, Perplexity, and Google AI Mode the same 10 questions a local diner would ask, and log whether you appear and which source each answer cites. When a cited source is wrong or stale, fix that source. The spreadsheet, query list, and tools are in the tracking post.

What does not work?

Keyword stuffing, fake or paid reviews, blocking the search crawlers to "protect" your menu, and paying an agency that promises guaranteed AI placement do not work, and the first two can get your listing suspended or draw a federal penalty.

Keyword stuffing. Google's spam policies name the patterns small sites fall into: "blocks of text that list cities and regions that a web page is trying to rank for" and "lists of phone numbers without substantial added value." Offending sites "may rank lower in results or not appear in results at all," and the AI features draw on the same index.

Fake, paid, or filtered reviews. Google's Maps content policy prohibits content "not based on a real experience," reviews "paid for, directly or in kind," incentives such as discounts for a review, and any attempt to "discourage or prohibit negative reviews, or selectively solicit positive reviews." The FTC's final rule on fake reviews (August 14, 2024) bans fake reviews including AI-generated ones, paying for reviews, undisclosed insider reviews, and review suppression, with civil penalties.

"Guaranteed AI placement" services. OpenAI says placement is not guaranteed and Google says no special optimizations exist. A vendor promising a guaranteed spot is doing the seven steps above at a markup, or nothing measurable.

Files nobody asked for. An llms.txt file will not hurt, but Google says you do not need "AI text files" for its AI features, and the OpenAI and Perplexity crawler pages above do not mention one.

Restaurant JSON-LD example you can copy

Paste this inside a script tag with type application/ld+json on your homepage and menu page, replace every placeholder, and check it with Google's Rich Results Test. Property names and the hours format follow Google's local business documentation.

{
  "@context": "https://schema.org",
  "@type": "Restaurant",
  "name": "Example Taqueria",
  "url": "https://www.example.com",
  "telephone": "+1-555-555-0100",
  "priceRange": "$$",
  "servesCuisine": "Mexican",
  "menu": "https://www.example.com/menu",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "123 Main Street",
    "addressLocality": "Springfield",
    "addressRegion": "IL",
    "postalCode": "62701",
    "addressCountry": "US"
  },
  "geo": {
    "@type": "GeoCoordinates",
    "latitude": 39.79842,
    "longitude": -89.64460
  },
  "openingHoursSpecification": [
    {
      "@type": "OpeningHoursSpecification",
      "dayOfWeek": ["Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"],
      "opens": "11:00",
      "closes": "21:00"
    }
  ]
}

To include the menu itself, the schema.org Menu type nests hasMenuSection and hasMenuItem, each item carrying an Offer with price and priceCurrency plus a suitableForDiet value. Leave aggregateRating out unless it comes from real reviews collected on your site; an invented rating is the markup-to-page mismatch Google tells you to avoid.

Where DirectOrders fits

DirectOrders gives an independent restaurant a branded, commission-free ordering website on its own domain, so the menu lives as pages an assistant can crawl and a Business Profile can link to, not behind a marketplace URL. Menu Brain adds nutrition and allergen clarity to each item, the detail behind "do you have gluten-free options." The discovery feature covers the AI search side, and the AI tools guide compares the options. Pro is $249 a month with a 14-day free trial, and none of the seven steps require it.

Sources

All pages checked September 2026.

• OpenAI Help Center: Searching the web with ChatGPT
• OpenAI: Overview of OpenAI crawlers
• OpenAI: Introducing ChatGPT search (October 31, 2024)
• Perplexity: Perplexity crawlers
• Bing Webmaster Tools: Which crawlers does Bing use?
• Google Search Central: AI features and your website
• Google Search Central: Google's common crawlers
• Google Search Central: Create and submit a robots.txt file
• Google Search Central: Local business structured data
• Google Search Central: Spam policies
• Google Business Profile guidelines
• Google Business Profile: Food ordering links
• Google Maps user-contributed content policy
• FTC: Final rule banning fake reviews and testimonials (August 14, 2024)
• Yelp content guidelines and Yelp for Business
• Tripadvisor: Claim your free business listing
• schema.org: Restaurant and schema.org: Menu

Bottom line

ChatGPT, Perplexity, and Google's AI features all read the open web, and their documentation says so: let OAI-SearchBot, PerplexityBot, and Googlebot crawl you, and there is no other gate. What decides whether you are recommended is what they find: a menu in real text with prices, structured data and a Business Profile that match it, listings and reviews on the platforms they name, and pages that answer what diners ask. Do the seven steps once, re-check the table monthly, and let the tracking log tell you which source to fix next.