// sector · Travel & leisure
A traveller gives ChatGPT their dates, budget, the kids' ages and what they never want to repeat, then asks where to go. Your offer can enter that plan. And what comes out of it can be measured all the way to the booking.
we're off for 10 days in August with two kids aged 4 and 7, €3,000 all-in, no more than a 3-hour flight, where should we go?
With kids that age and a short flight, the package outweighs the country: a kids' club by age group, included transfers and connecting rooms change a holiday more than the destination does. Before comparing prices, check the minimum age for the kids' club, the airport-to-hotel distance, and what "all-inclusive" actually leaves out.
Family clubs in August
Kids' club from age 4, flights in, live stock.
your-tour-operator.com
// why this sector
Nobody types "hotel Lisbon" into ChatGPT; they tell it about a trip. They say who's going, when, on what budget and what to avoid, so everything an agent would ask over the phone lands in the first message.
Dates, length, group make-up, budget, constraints such as a pushchair, mobility or diet: an agent rebuilds all of that question by question, whereas ChatGPT receives it in one block. That block is what the context hint learns to recognise.
Many conversations open with "where should we go" and only a period and a budget: at that stage an offer can still enter the plan, whereas a search on a city name comes well after the decision.
School holidays, long weekends, late bookings: planning peaks come before departure peaks. So ad groups open when people plan and close when stock is sold. Departure comes much later.
// intents
Typical prompt
"Where to go for a week in October, two of us, €1,500, somewhere sunny, no night flight?"
What the ad has to do
Enter the plan before the destination is set: the ad names the situation (October, couple, sun), not a list of countries.
Typical prompt
"All-inclusive club or a rental with a car for a family of five in Sicily?"
What the ad has to do
Carry one verifiable difference, what is included and what isn't, with one argument per ad.
Typical prompt
"Three days in Porto in November with a teenager, what to do besides the old town?"
What the ad has to do
Sell the activity or ticket just as the itinerary is built. The landing page shows time slots, not a brochure.
Typical prompt
"Will February ski packages get cheaper if I wait, or should I book now?"
What the ad has to do
Answer the window question with remaining availability, cancellation terms and the displayed price, promising nothing about how prices will move.
// context hints
A traveller's hint holds a period, a budget and a group, rarely a firm destination or a geographic zone. Here are three ad groups for a family-holiday operator:
Generate your context hintsIllustrative excerpt. The real structure is built from your search terms, sales calls and support tickets.
AG · Family holidays
AG · Last minute
AG · Sports trips
// scenarios
Three examples that tie a conversation to an ad, a page and a metric. They are not delivery promises.
// measurement
OpenAI's report gives impressions, clicks, spend, CTR, average CPC and average CPM, then stops. Yet a booking often lands days later, sometimes from another device, so the click alone tells you very little. So the OpenAI Pixel and the Conversions API are wired to your booking events, and the account is read against the Hiints Index.
Confirmed bookings
The reference event: the paid booking or the accepted quote, never the click or the listing viewed.
Availability requests
For an independent hotel or a rental, this is often the real conversion. We measure it as a lead, then follow its confirmation.
Average basket per booking
A family holiday and a midweek night can't carry the same CPC: cost per booking is read against the value of the stay.
Click → booking delay
In travel it's counted in days. The attribution window has to cover it, or the channel looks less profitable than it is.
Cancellations
A cancelled booking isn't a conversion. We reconcile cancellations with the PMS or CRM to correct the number.
Your season vs the Hiints Index
Your CPC and CTR set against comparable travel accounts. OpenAI publishes no benchmark, so this reference doesn't come from OpenAI.
// honest qualification
Yes, clearly
You sell stays, nights, rentals or activities with an average basket above €100, the booking or quote request happens online, and you can show prices and availability on the landing page.
Yes, with one caveat
Your offer depends on a single date or a single place, with a very short booking window: hints describe situations, not zones, and volume may be lacking. If you also sell travel insurance, that category falls under OpenAI's advertising policies and its eligibility is checked before any launch.
Not yet
With no online booking, and sales that end on the phone leaving no trace, there is nothing to measure. A monthly media budget under €3,000 doesn't give enough volume to learn either. In both cases, we tell you.
Free setup, done in under 48h. Then 15% of ad spend under €5,000 of monthly media, 10% above. See pricing →
// questions
Yes, provided your guests can book or request availability directly on your site: the ad then points to the availability page, never the homepage. If all your sales go through a third-party platform, measurement becomes impossible, and we'll say so rather than launch anyway.
The delivery country is set at campaign level. Context hints are not geographic rules: they describe a traveller and their situation, and they guide matching without guaranteeing it. A hint can describe someone looking for a family week by the sea; it can't order "only show in Paris".
When your customers plan, not when they leave. Your booking history shows the window. Ad groups open before the peak and get cut when stock is sold.
The confirmed booking when it happens online, the quote or availability request when the sale closes in a conversation. Both are tied to the OpenAI Pixel and the Conversions API, and cancellations are reconciled so a sale that didn't happen is never counted.
No. "cheap hotel Lisbon" is a query, not a situation. Your search terms help you understand demand, but the structure gets rewritten around intent families: period, budget, group make-up, type of stay.
Three things set it: your average basket, your margin per booking and the volume a single season lets you observe. Setup is free, and the fee follows your media, at 15% while it stays under €5,000 a month and 10% above. Under €3,000 of monthly media, we'd rather not have you test.
// keep reading
Context Hints in ChatGPT Ads: Definition, Format and Examples
What a context hint is according to OpenAI, where it is set, the who/when/what format, five good-versus-bad examples and the mistakes to avoid.
Read →How-toHow to Structure a ChatGPT Ads Account: Campaigns to Ads
One campaign per objective, one ad group per intent family, one ad per angle. Naming rules, intent families and a complete example for a B2B SaaS.
Read →GuideHow Much Does ChatGPT Ads Cost in 2026: CPC, CPM and Minimum Budget
Observed CPC of $3 to $5, CPM of $25 to $60, $25 per day minimum per campaign, no overall minimum since May 2026. What sets each cost line.
Read →// shall we talk?
Thirty minutes, live, and your category is placed in the Hiints Index. If travel isn't the right ground for your offer, you'll know before the call ends.
Book a call30 minutes · no commitment · zero pitch
// other sectors
