Try an experiment. Open ChatGPT and type: "What do you know about Hotel [name] in [town]?" Then ask: "Recommend three hotels in [town] for a family with kids." Does your hotel appear? Is the information correct?
The first time many hoteliers run this test, they discover surprising things: amenities they don't have, renovations that never happened, made-up prices — or simply no mention at all. In this article we look at how to check whether AI assistants know your hotel, with a simple, repeatable method.
Which assistants to check
The most used for travel planning are:
- ChatGPT (OpenAI);
- Gemini and Google's AI Overviews / AI Mode;
- Perplexity;
- Copilot (Microsoft);
- Claude (Anthropic).
Some answer mainly from what they've learned; others search the web in real time and cite sources. That's why answers can differ a lot.
Test prompts
Prepare a fixed list of questions to repeat every few months in the same way.
Questions about your hotel
- "What do you know about Hotel [name], [town]?"
- "Does Hotel [name] in [town] have parking? Does it accept pets?"
- "What's breakfast like at Hotel [name]?"
- "What are the pros and cons of Hotel [name]?"
Recommendation questions (the most important)
- "Recommend a hotel in [town] for a family with two kids."
- "Quiet hotel near [point of interest] with parking."
- "Where to stay in [town] for [event/trade fair]?"
- "Good-value hotel in [town]."
Adapt the questions to your real target guests. If cyclists are your strength, ask about cyclist-friendly hotels.
In your guests' languages
Repeat the main questions in English, German or the languages of your markets. Answers can change a lot.
How to record the results
Use a simple table:
| Question | Assistant | Does my hotel appear? | Correct information? | Sources cited | Competitors cited |
|---|---|---|---|---|---|
| Family hotel in… | ChatGPT | No | – | – | Hotel A, Hotel B |
| What do you know about Hotel… | Perplexity | Yes | Wrong parking info | Booking.com, Tripadvisor | – |
The "sources cited" and "competitors cited" columns are valuable: they show where AI gets its information and who's winning.
Tips for reliable tests
- Use a clean session (logged out or a new chat) to reduce the effect of previous conversations.
- Repeat the same question several times: AI answers vary.
- Note the date and tool version.
- Don't draw conclusions from a single test: look at trends over time.
What to do with the results
If your hotel doesn't appear
Online sources probably don't describe clearly enough who you are and who you're right for. Work on your website, Google profile, content and reviews. See GEO for hotels: what it means to be found in AI answers.
If it appears with wrong information
Find the source of the error (often an OTA or directory with old data) and correct it. See AI describes your hotel incorrectly: what can you do?.
If it appears, but never for your target
AI knows you but doesn't associate you with the right needs. Make your positioning explicit on your website, OTAs and Google profile ("family hotel", "ideal for cyclists").
Frequently asked questions
Can I ask ChatGPT to correct information about my hotel?
Not permanently, directly. Corrections come from improving online sources: website, profiles, OTAs, reviews. Some tools offer feedback features, but the main route is the quality of your sources.
How often should I repeat the test?
Every 2–3 months is a good rhythm — and always after major changes (new website, renovation, new services).
Are there tools to automate these tests?
Yes, tools that monitor brand presence in AI answers are emerging. They're useful, but manual tests remain a great starting point. See your hotel's visibility in AI answers: what you can and can't measure.
Want to know what AI says about your hotel?
With our free marketing audit we test how the main AI assistants describe and recommend your hotel and show you what to improve to appear in the right answers.



