ChatGPT Ads · Hotels and accommodation
ChatGPT Ads for Hotels and Accommodation
Travel planning rarely starts with a hotel name, it starts with a situation: a long weekend, a wedding nearby, a trip with the dog. Exactly these conversations happen in ChatGPT, and travel and experiences are among the categories OpenAI started with. Your ad sits where a vague idea is turning into a plan.
Travel and experiences are among the verticals OpenAI started with according to its ad policies (as of 31 August 2026), and hotels and holiday rentals sit right inside that category.
Conversation starters
How people ask in ChatGPT
Not search terms, conversations. This is roughly how the questions sound next to whose answer your ad should appear.
Where can I go for three days in November without a long drive?
Which hotel in the region is good for a weekend with a dog?
I am looking for a place where all our wedding guests can stay.
What do we do with the kids if it rains the whole week?
Is it worth booking directly with the hotel instead of a portal?
Context hints
Context hints that match
Context hints describe the situation in which your offer helps. OpenAI matches them against the running conversation, not against a keyword.
Couples planning a short weekend in the region who want a place with a sauna and late arrival
Families with small children who need a bad-weather programme and a room with a connecting door
Wedding parties looking for a room allocation within walking distance of the venue
Business travellers with an early departure who need parking and breakfast from six
Dog owners looking for a stay without surprise surcharges and with somewhere to walk nearby
Examples following the OpenAI best practice, not exact match keywords.
Setup
What the campaign starts with
Campaign objective
If your booking engine measures cleanly, order_created is the target event for a completed direct booking. You still start with clicks (CPC), and only once the event fires regularly do you create a new campaign with the conversions objective (oCPC) on it.
Landing page
The ad points at the concrete offer with room type, dates and price, not at a homepage with an image slider. If your booking engine sits on another domain, UTM parameters and the oppref click reference have to survive the handover, and OAI-AdsBot and OAI-SearchBot must not be blocked.
From my practice
Three mistakes that tend to cost money
Advertising only the brand and the homepage. From experience in auction channels, it is the occasion that converts, not the hotel name.
A booking engine on a third-party domain without forwarded parameters. From my practice with auction channels, transferred to ChatGPT Ads: the conversion is then missing from your report and you steer blind.
Prices in the ad that no longer apply on the page. OpenAI requires consistency through to the landing page, and in my experience guests bounce immediately.
Pricing
What the start costs
Account, pixel, context hint map and the first campaign, live in 7 days.
3 months start phase, cancellable monthly after that.
Review of an existing account, including a list of actions.
Frequently asked questions
Am I competing against the booking portals here?
You compete for the same conversation, but with a different promise. Portals win on selection, you win on proximity to the occasion, the direct rate and services that are invisible on a portal. That is exactly what belongs in your context hints and ad copy.
Which occasions work best as context hints?
From experience in auction channels, the ones that describe a full situation: a weekend for two with a late arrival, wedding guests within walking distance, a trip with a dog. OpenAI recommends exactly this level of detail instead of a keyword, while guaranteeing no specific conversations.
Can I re-engage guests who have stayed before?
For campaigns targeting the EEA or Switzerland OpenAI does not support custom audiences, and ads there are not personalized at launch. Re-engagement therefore stays with your own channels rather than ChatGPT Ads.
When can I tell whether the test is paying off?
OpenAI treats the daily budget as an average across seven days, so individual days can run higher. Judging after two days means judging too early: define a time window and a clear stop criterion before launch.