AI Visibility Research
66% of restaurants are invisible to AI assistants. ChatGPT relies on TripAdvisor, Yelp, and Google reviews to recommend dining options — restaurants without strong directory profiles and menu schema are skipped entirely.
Check your business66%
Invisible to ChatGPT
restaurants
Thousands
Businesses analysed
34%
AI-visible
Can be found by AI assistants
2-4 weeks
Time to fix
With structured data + directories
The test
ChatGPT recommends restaurants primarily from TripAdvisor, Yelp, and food blogs. Only 34% of recommendations link to actual restaurant websites. Restaurants with text menus and schema are 3x more likely to be directly cited.
Each AI system uses different data sources and citation patterns.
Source: Bing index
ChatGPT recommends restaurants primarily from TripAdvisor, Yelp, and food blogs. Only 34% of recommendations link to actual restaurant websites. Restaurants with text menus and schema are 3x more likely to be directly cited.
Source: Live web crawl
Perplexity cites restaurant websites more often (52%) because it follows live links from Google Maps results. Recently updated sites with clear location data perform best.
Source: Google index + Maps
Google AI Overviews for restaurant queries pull directly from Google Maps and GBP. Review count, recency, and photos drive which restaurants appear in AI-generated summaries.
Restaurants businesses face specific challenges that prevent AI assistants from finding and recommending them.
Restaurant menus are frequently embedded as images or PDFs — AI cannot read 'wood-fired pizza' from a JPEG. Search engines index text, not photos of chalkboards.
Many restaurants have no website at all, relying entirely on social media. ChatGPT cannot cite an Instagram page as a restaurant recommendation.
Review platforms (TripAdvisor, Yelp, Google) dominate food recommendations in AI. Restaurants without strong review profiles across multiple platforms get overlooked.
Opening hours, cuisine type, and price range are often missing or inconsistent across different directories, causing AI to skip unreliable listings.
What the visible businesses do differently
Text-based menus on their website with Menu schema markup (itemListElement for each dish, price, description) • 100+ reviews on Google with an average rating above 4.2 — AI heavily weights review volume and recency • Consistent listings across TripAdvisor, Yelp, OpenTable, and Google with matching hours, phone, and address • Regular GBP posts showing daily specials, events, and seasonal menus — signals freshness to AI
Schema types that matter: Restaurant, Menu, FoodEstablishment, LocalBusiness
Replace image/PDF menus with HTML text. Add Menu schema to each item with name, description, price, and dietary info (vegan, gluten-free). This is what AI reads when someone asks 'restaurants with vegan options near me'.
Google reviews matter most, but TripAdvisor and Yelp are what ChatGPT cross-references. Ask every happy customer. QR codes on tables, links on receipts. Target 100+ reviews on Google.
Include cuisine type, price range ($ to $$$$), hours, reservation links, and accepted payment methods. AI uses this to match 'cheap Italian near me' queries.
TripAdvisor, Yelp, OpenTable, Zomato, TheFork, and Google. Ensure hours, phone, address, and menu are identical everywhere. AI penalises inconsistency.
Share daily specials, events, and seasonal menus. GBP posts signal to Google and AI that your restaurant is active and current — stale profiles get deprioritised.
Only about 34% of restaurants are visible to ChatGPT. The rest are invisible because they lack the structured data, directory presence, and content that AI assistants need to make recommendations. Our research shows that ChatGPT relies on Bing's index for 87% of its citations.
Most restaurants can achieve basic AI visibility within 2-4 weeks by adding structured data, claiming key directories, and submitting to Bing Webmaster Tools. Full optimisation (including review building and content creation) typically takes 2-3 months to show results.
Not directly. ChatGPT uses Bing's index, not Google's. However, the factors that help you rank on Google (structured data, quality content, reviews) also help on Bing. Our data shows that 87% of ChatGPT citations match Bing's top 10 results, while only 56% match Google's top 10.
It depends on the query. Google AI Overviews are strongest for local searches backed by Google Maps data. Perplexity crawls the web daily and often finds more niche results. ChatGPT relies on Bing's index and favours well-known directories.
restaurants should use Restaurant, Menu, FoodEstablishment, LocalBusiness schema types. The most impactful is adding your business details (name, address, phone, hours, services) in structured data that AI can read programmatically.
Run a free Flatline audit to check your visibility score. We'll tell you if ChatGPT, Perplexity, and Google AI can find your business — and what to fix if they can't.
Free SEO audit in 15 seconds
Run free audit