Trip.com sets out three years of contrasting AI use in travel

Nearly 60% of interactions now booking-related

Trip.com has taken stock of how TripGenie, its AI assistant launched in 2023, is being used. According to aggregated data, the tool is now deployed across the travel journey, though the pace and expectations vary by market. The company reports a marked rise in booking-related interactions, which now account for close to 60% of exchanges, and a shift from a functional helper to an end-to-end travel companion.

In short-haul, highly connected markets such as Hong Kong, Singapore and Malaysia, travellers are more likely to lean on AI in situ to arbitrate real-time choices — before booking a hotel, to navigate attractions or to find inspiration on the spot. In Europe and North America, usage typically happens earlier in the funnel. In Germany and the UK in particular, travellers consult the assistant weeks before departure, chiefly to compare flights and accommodation and to lock in decisions.

A surge in multimodal usage

Decision speeds also diverge. In South Korea and Taiwan, hotel bookings are often finalised just days before departure, with AI helping decide on location and nearby services. Japan is an exception in Asia, with accommodation booked further in advance. In Italy, France and Spain, behaviours look more spontaneous.

The search for certainty underpins these dynamics. TripGenie is used to verify baggage policies or lounge access on the air side, and to check whether a property meets specific needs on the accommodation side. According to Trip.com, hotel comparison tools cut the number of clicks required by 80% and are associated with a 45% uplift in seven-day revisit rates. Pre- and post-sales queries make up roughly a quarter of interactions, suggesting the AI is handling everyday, practical cases from itinerary changes to in-destination questions.

The report also highlights a rise in multimodal usage, beyond text. Travellers are increasingly sharing images — menus, signs, rooms and transport information — to obtain contextual answers. These users post a seven-day revisit rate twice the platform average, a signal of stronger satisfaction and trust. More broadly, Trip.com stresses that the assistant supports local behaviours rather than homogenising them, adapting to practices in each market.

The company says AI-assisted order volume has risen by around 400% year on year. For Trip.com, these signals confirm a shift from one-off search to cross-journey assistance, from inspiration through to service interactions.