What AI says about hotels
ChatGPT, Gemini and Google's AI Overviews answer questions about hotels every day, and the people asking rarely double-check. Here is what they get wrong about a hotel, why, and how to find out what yours says.
What is at stake for hotels
Hotel errors do not just lose bookings, they create bad reviews. A guest who arrives expecting free parking because an engine said so is angry at you, not at the engine, and writes about it.
The awkward part is that none of this shows up as a problem you can see. There is no bounced call and no complaint — the customer asked, got an answer, and went somewhere else. The first sign is usually a quiet month you cannot account for.
What customers actually ask
These are the questions being put to engines about hotels, and they are worth reading closely: none of them is your business name on its own. People ask about a need, and the engine decides who fits.
- “Does [hotel] have free parking?”
- “What time is check-in and check-out at [hotel]?”
- “Does [hotel] allow dogs?”
- “Is breakfast included at [hotel]?”
- “Best hotel in [town] for families?”
Ask each of those about your own hotel, from a private window, and read what comes back. That is the audit — there is not a more sophisticated version of it.
The errors that specifically bite hotels
Every trade has its own failure modes, and generic advice misses them. These are the ones that recur for hotels:
- Parking described as free when it is paid, or the reverse
- Pet policy out of date, which produces arguments at the desk
- Amenities listed that have closed — pool, restaurant, gym
- Check-in times wrong after an operational change
The listings that matter here
Google Business Profile is the foundation for everyone, but the listings that carry weight after it differ sharply by trade. For hotels these are the ones worth auditing:
- Google Business Profile
- Booking.com and Expedia, which engines read as authoritative on amenities
- TripAdvisor
- Your brand’s own property page, if you are part of a group
- Apple Maps
Reviews for hotels accumulate mainly on Google, TripAdvisor, Booking.com, Expedia — and the text of those reviews is being read, not just the star average. A model asked for the best option for a specific need can only name you if somebody wrote about that specific need.
The one thing most hotels get wrong
Audit your OTA listings as carefully as your own site. Booking.com and Expedia amenity checkboxes are treated as highly authoritative on exactly the questions guests ask most, they are frequently filled in once at launch and never revisited, and a stale checkbox there will override the correct information on your own website.
How to check yours
- Open a private window. In ChatGPT, turn memory off — otherwise it may be answering from something you told it months ago.
- Ask the five questions above, using your full street address so you are not reading an answer about a business with a similar name.
- For anything wrong, ask “what sources did you use?” Fix the page it names. If it cannot name one, the detail was invented and the fix is making the correct fact easier to find.
- Paste the answers into a document with today’s date. A single audit tells you about today; what you need to know is whether things improve after you change something.
- Check again in a month. Corrections propagate on their own schedule.
The longer version of this process, with the exact prompts, is in how to check what AI says about your business.
If you would rather not do it by hand
Once a month by hand is fine for one location that rarely changes. It stops being practical when you have several, or when you need to know the week something breaks rather than the month.
AI Business Monitor runs these questions weekly for your hotel, stores every answer so you can see what changed, and emails you when something moves — with the source behind each finding. It cannot change what an engine says, and neither can anyone else; what it does is tell you what is being said.
Related reading
- AI says my business is closed — the most expensive error there is, and the one to fix first.
- Wrong hours, phone or address — tracing each error back to its source.
- Getting recommended by AI — what actually influences whether you are named.
Common questions
How do I find out what ChatGPT says about my hotel?
Ask it directly, in a private window with chat memory turned off so it is not answering from things you told it before: "What do you know about [name] at [full street address]? Include hours, services and contact details." Then ask "what sources did you use?" — the page it names is usually the thing that needs fixing.
What does AI most often get wrong about hotels?
Parking described as free when it is paid, or the reverse. Pet policy out of date, which produces arguments at the desk. Both come from a source that was correct once and was never updated, which is why finding the source matters more than correcting the engine.
Can I make an AI engine correct itself?
Not directly — there is no listing to edit and no correction process, and nothing typed into a chat window teaches a model anything about your business. What works is fixing the underlying sources it reads, then waiting. Google Business Profile changes often show up within days; ChatGPT can take weeks.
Which listings matter most for hotels?
Google Business Profile first, always. After that the ones specific to this trade: Booking.com and Expedia, which engines read as authoritative on amenities, TripAdvisor, Your brand’s own property page, if you are part of a group. Trade-specific directories carry more weight for these questions than general ones, and they are the listings owners most often forget exist.
Check it every week automatically
Doing this by hand once tells you about today. Things change: hours get edited, reviews land, a competitor opens. We check weekly and email you when something moves, with the source behind every finding.