How ChatGPT decides which hotels to name
When a traveller asks "Which boutique hotel in Lisbon is good for a quiet weekend for two?", ChatGPT draws on two things. The first is its training data: a snapshot of the web that can be a year or more old. The second is live web search. For a travel question it will usually search, read a handful of pages and write a short answer that names three to five hotels.
ChatGPT writes that answer from the pages it has just read. Your hotel gets named when two things are true: it appears in those pages, and what is written about it matches the details in the question.
Those details matter more than they do in a classic search. People type "hotel Lisbon" into Google. They ask an AI platform for a quiet hotel, walkable to the old town, with parking, under €200 a night. A hotel that is only ever described as "charming and centrally located" gives ChatGPT nothing to match against that question.
Where the information comes from
ChatGPT rarely relies on one source for a hotel. An answer is typically stitched together from several of these:
- Your own website. The most detailed source for facts such as room types, facilities, policies and location.
- Booking and review platforms. Booking.com, Tripadvisor, Expedia and Google reviews supply ratings and the words guests use to describe a stay.
- Editorial coverage. Travel guides, newspapers, local blogs and "best hotels in" lists. These carry a lot of weight, because a "best hotels for" question matches a "best hotels for" article almost word for word.
- Destination and partner sites. Tourism boards, hotel associations, wedding and event venues, tour operators.
The practical consequence: a perfect website is not enough on its own. If the pages ChatGPT reads about your destination never mention you, your own site may not be consulted at all.
Five things you can do
1. Let the crawlers in
OpenAI runs separate crawlers. OAI-SearchBot finds pages for ChatGPT's search answers, ChatGPT-User fetches a page when a user asks about it, and GPTBot collects training data. Open yourhotel.com/robots.txt and check that none of them is disallowed. Then ask whoever manages your hosting whether a firewall or bot protection service blocks them. This happens more often than hoteliers expect, and it is invisible from the outside.
2. State the facts in plain text
Write out what a guest would ask at the front desk: how far you are from the station and the main sights, which room types you have, whether there is parking and what it costs, breakfast times, pet policy, check-in and check-out, accessibility. Put it in text on the page. Facts that live only in images, PDFs or a booking widget are easy for a crawler to miss, and many AI crawlers do not run JavaScript, so content that loads through a script may never be seen.
3. Say who the hotel is for
"A 24-room adults-only hotel, five minutes on foot from the cathedral, suited to couples and solo travellers who want quiet evenings" gives an AI platform something to match. "A unique experience in the heart of the city" does not. Be specific about the type of stay you are good at, and about the ones you are not.
4. Keep your details the same everywhere
Compare your website with your Google Business Profile, Booking.com and Tripadvisor listings. The hotel name, address, phone number, star rating, facilities and room count should agree. When sources contradict each other, an AI platform either repeats the wrong one or leaves the detail out.
5. Earn mentions where AI reads
Search for the questions your guests ask and look at which articles and lists come up. Those are the pages ChatGPT is likely to read. Ask your tourism board to complete your listing, give local writers a reason to include you, and make sure partner sites link to you with an accurate description.
See where your hotel stands today
Before you change anything, find out what AI platforms already say about you. The free myhotail scan takes your hotel name and city and shows whether ChatGPT, Gemini and Perplexity recommend you today. No card needed.
For the full picture, the Audit puts everything AI sees about your hotel into one document, delivered by email within 24 hours. The Dashboard then tracks your visibility every month across all five AI platforms and tells you exactly what to fix next.
Talk it through with us
Book a 20-minute call and we will show you what AI platforms say about your hotel today and what we would fix first.