The news. Phocuswright surveyed more than 1,550 leisure travellers across the UK, France and Germany. As TravelDailyNews reports it, fewer than one in four in any of the three markets completely or mostly trust AI-generated travel results. They want AI to organise options and explain trade-offs, and they still expect to verify the information, compare alternatives and override the recommendation.
The same survey looked at personal data. In the UK, 45% accepted a company storing travel documents and identity details for a faster check-in. Using search, browsing and purchase history to personalise offers drew 37% in the UK, 31% in France and 28% in Germany. Phocuswright's managing director put it this way: ask for data only when the traveller can see what they get back, and "show your work when AI makes a recommendation."
I read this write-up, not the report itself, so treat these as the headline numbers.
This is how clients already behave
Nobody who has worked an agency desk should be surprised. A client asks for options, wants to know why one costs more, and then goes off to check. The checking does not stop when the client says yes.
In the booking record from the desks I ran, 25.3% of more than 34,500 decided bookings ended in a cancellation. A confirmed booking was often still a question in the client's head.
What to do with it
So the useful job for AI on a travel desk is the one these travellers describe: lay out the options, say what each one costs you, and leave the decision visible and reversible. A tool that hides its reasoning asks for a trust that fewer than one in four are offering. A person who can explain the trade-off still earns it.
When a client overrules your recommendation, do you know which part of it they stopped believing?
Source: TravelDailyNews International, European travellers favour sharing passport details over browsing history, Phocuswright finds, reporting Phocuswright's study "Europe's Trust Economy".