How to get your hotel recommended by ChatGPT, Gemini and Perplexity
“Which hotel should I pick in Queenstown for skiing?” A traveller planning a trip increasingly asks an AI assistant, then books one of the four or five names it gives. Here is where those names come from, based on our measurements, and how to get your property among them.
By the PickableAI team · Updated 8 October 2026 · 2 min read
What our measurements show
In our ranking of hotels in a ski resort (28 September 2026), AI describes each hotel in one very concrete sentence: “ski-in/ski-out, next to the gondola”, “spa”, “Alpine style with modern comfort”, “apartments with hotel services”. It does not recommend “a good hotel”: it recommends a hotel for a need.
Second finding: to answer, AI combines several kinds of sources, not just booking sites. The list is below.
Where AI gets its information about hotels
In that ranking, the most cited sources were:
- the resort's tourist office website;
- a hotel group's website, for chain properties;
- destination guides written for travellers;
- a booking site (Booking.com);
- hotel collections (The Hotel Guru, Small Luxury Hotels of the World);
- YouTube, for filmed tours.
For an independent hotel the conclusion is encouraging: the tourist office website comes first, and it is open to every property in the destination.
The 5 actions, in order
1. See which need you are named for, or not
Ask several questions, not just one: “family hotel in Whistler”, “hotel with a spa in Banff”, “cheap hotel near the slopes”. The free check asks four standard questions; the plans let you add your own.
2. A complete listing on your tourist office website
It is the most cited source in our measurements. Check your listing: category, up-to-date description, facilities, recent photos, link to your website. A listing created several years ago deserves a second look.
3. Write your strengths as answers
AI repeats precise facts. On your website, as text: distance to the lifts or the station in minutes, spa and pool, family rooms, parking, shuttle, pets allowed, breakfast included, year of renovation. One page per need beats a single “The hotel” page: family stays, seminars, wellness weekends.
4. The same information everywhere
Exact name, star rating, address and facilities identical on your website, your Google Business Profile, booking sites and the tourist office. If AI reads “3 stars” in one place and “4 stars” in another, it no longer knows which one to repeat.
5. A website crawlers can read
Many hotel websites show rooms and rates inside a script-loaded booking widget: AI crawlers see nothing there. Keep a text description of each room type, then test with Your site as AI sees it. Add the structured data block (type Hotel) with the generator.
Common mistakes
- Superlatives instead of facts: “an exceptional setting” answers no question; “200 metres from the gondola” answers many.
- Leaving everything to booking sites: they describe your hotel in their words, not yours.
- Seasonal pages taken down out of season: keep them online all year, AI answers winter questions in summer too.
Frequently asked questions
Which sources does AI use to recommend a hotel?
In our ranking of hotels in a ski resort (28 September 2026), the most cited sources were the tourist office website, a hotel group's website, destination guides, a booking site, hotel collections and YouTube.
Can an independent hotel be recommended alongside the chains?
Yes. In our measurements the most cited source is the tourist office website, which is open to every property. In the answers we collected, independent hotels appear next to chain properties.
What does AI remember about a hotel?
Precise facts tied to a need: access to the slopes, spa, style, type of accommodation. In the answers we collected, each hotel is described in one concrete sentence, never just as “a good hotel”.
Does AI recommend you?
The free check asks AI assistants your customers' questions (“best {trade} in {city}”) and shows your position, or who is named instead of you.
Check my business