True to his financial roots, Adrian Currie, former Booking Holdings executive, opened his panel on “OTAs: The Next Frontier”, with numbers to illustrate the size of the four companies on his panel. Booking.com, Expedia Group, Trip.com Group and, in spirit, Katrina Barry’s former employer Webjet collectively process close to half a trillion US dollars in travel transactions a year, roughly two and a half billion room nights, across hundreds of millions of customers.
The big question – how much of that machinery AI is actually about to rewire.
From numbers, he then switched to personal stories with each panellist recounting how they got to where they are today.
Todd Lacey of Booking.com, a Kiwi based in Sydney, first came to Queenstown around age 10 on a ski trip during the original Lord of the Rings era and noted that LOTR-themed tours and the TSS Earnslaw steamship cruise remain among the platform’s most-booked regional experiences to this day, more than two decades later.
Darren Karshagen of Expedia, originally from South Africa and also Sydney-based, first visited in 2019 while running the holiday rental side of the business, and was struck less by the scenery than by the range, a destination equally set up for luxury travellers and backpackers.
Katrina Barry, now a trustee of Technology Queenstown after stepping down as Webjet’s CEO, first came on a family camping trip as a kid from Auckland, and described the unpaid trustee role as something she took on because the destination and the New Zealand tech and travel ecosystem behind it genuinely matters to her.
As for Ru Yi of Trip.com, based in Singapore, her first visit to Queenstown was exactly a decade ago, when her daughter turned four on the same day the Earnslaw turned 104 – a hundred years apart, almost to the day.
Four mission statements that all sound the same
Currie then got into their mission statements, stating that they were nearly indistinguishable. He read each one back and asked for the real differentiator.
Lacey’s answer for Booking.com centred on what the company calls the Connected Trip: stitching accommodation, flights, ground transport and attractions together so a traveller isn’t managing five separate bookings, and so the platform can intervene proactively when something breaks – his example was a cancelled or delayed flight into Queenstown automatically triggering a later taxi pickup and a new restaurant suggestion, adjusted for dietary preference, before the traveller even realises they need it.
Karshagen’s answer leaned into Expedia’s B2B ambitions as a genuine growth engine in their own right – partnerships like the one with Uber, and API-level integrations covering car rental, ground transport and insurance, aimed at making Expedia’s travel inventory available inside other companies’ ecosystems, not just its own branded apps.
Darren Karshagen said Expedia now generates AI-drafted, property-customised responses to routine guest questions that hosts can approve and send with one click and more than 75% of those drafts are accepted without any edits
Currie used the word “dominant” to describe Trip.com’s position in China, and Ru Yi pushed back gently on that. Domestic China, she said, now contributes less than half of the group’s total production – the rest comes from Trip.com’s international sites across other markets. The current focus is twofold: continued aggressive expansion across APAC, and genuine localisation rather than translation – even markets that nominally share a language, like Singapore and Malaysia, have different customer preferences that require different product experiences on each local site.
Barry, drawing on her Webjet background, said the reason all four mission statements sound alike isn’t laziness, it’s that every OTA is chasing the same underlying shift in customer behaviour – from flight-led or accommodation-led booking toward genuinely travel-led planning, where the customer wants someone to assemble the whole trip rather than compare seventeen separate sites themselves.
Webjet built “joy” into its own mission deliberately, she said, because planning should be enjoyable but has become genuinely complex and because trust matters just as much as ease: travellers want a “travel companion” who’s still answerable if a booking falls over, not just a comparison engine that disappears the moment something goes wrong.
Where AI is actually doing work today
Pressed for concrete, operational (not theoretical) AI wins, Karshagen pointed to guest communications: Expedia now generates AI-drafted, property-customised responses to routine guest questions that hosts can approve and send with one click and more than 75% of those drafts are accepted without any edits, freeing accommodation partners’ time for higher-value work.
Ru Yi’s example was Trip.com’s Trip Planner tool: a traveller enters dates, destination, party size and interests, and the system assembles a customised itinerary with bookings attached — work that used to be done manually, one itinerary at a time. Roughly 18% of itinerary planning volume is now handled this way, freeing staff to focus on the more complicated cases that still need a human.
Ru Yi said roughly 18% of itinerary planning volume is now handled by Trip.com’s AI Trip Planner, freeing staff to focus on the more complicated cases that still need a human.
A second Expedia contribution focused on the supplier side: localised, digestible, increasingly chat-based dashboards that tell accommodation partners when their traveller is booking, what their profile looks like, and when to optimise pricing or availability – a faster, more curated alternative to the spreadsheets and static charts partners used to be handed.
Currie then turned the conversation toward the market’s own nerves about AI, noting that OTA share prices had a volatile start to the year – a sharp drop tied to broad AI-disruption anxiety, a recovery once several CEOs used earnings calls to reassure investors they had a credible AI position, and renewed jitters heading into the next reporting cycle.
He asked about Expedia’s own April “trust report,” which argued trust was the reason OTAs weren’t going away even as AI reshapes discovery. Karshagen’s answer: AI has become a genuinely excellent tool for planning, comparing and researching — an outstanding discovery layer. But the moment a traveller has to hand over payment details and personal information, trust becomes the deciding factor, and travellers consistently move to an established, trusted travel brand to actually transact, specifically because they want a real party to call if something needs to change.
The biggest opportunity, in his view, is combining the two – great AI-native discovery and content, fused directly with the trusted transaction layer – rather than treating them as separate systems.
Lacey brought the conversation back to New Zealand specifically, citing around 23,000 properties listed on Booking.com in the country, the vast majority small, independent, mum-and-dad operators without a strong SEO-optimised brand website – let alone, he joked, an AI assistant downloaded on their phone yet.
Booking.com’s value to that long tail, he argued, is absorbing exactly that complexity: automatically distributing small operators’ inventory across channels they’d otherwise have no way of reaching themselves.
On where AI ultimately takes customer behaviour, his answer was candidly noncommittal: nobody fully knows yet, so Booking Holdings’ strategy is simply to be present everywhere it might matter – built into its own platform, used internally, embedded with partners, present inside major LLMs, and backed by an active startup-acquisition strategy at the Booking Holdings level to keep new ideas flowing in rather than betting the business on a single guess.
Todd Lacey said Booking.com’s value to long tail is automatically distributing small operators’ inventory across channels they’d otherwise have no way of reaching themselves.
Fifty percent, 1.5%, and 94% invisible
Currie handed Barry three numbers and asked her to make sense of them: roughly half of travel planning is now reported to involve AI somewhere in the process; Expedia’s own last earnings report put AI-driven traffic at just 1.5%, albeit growing fast; and Cornell research reportedly found 94% of hotels are currently invisible to AI systems entirely.
Barry didn’t dispute any of it – she called the numbers believable, even if the gap between them looks contradictory on the surface. Her explanation: the industry spent 20 years optimising for search and paid acquisition and is only now beginning the equivalent work for AI discovery, moving in her words “from SEO to GEO to AEO.”
The reason inspiration skews so heavily AI-assisted while booking still overwhelmingly happens on established brands, she argued, comes back to trust and operational complexity that AI hasn’t solved yet – ticketing, refunds, and the years and genuine capital it takes to build brand equity that travellers are willing to bet a transaction on.
She expects AI to become the primary discovery layer over time but doesn’t see that displacing the trusted transaction layer any time soon.
Katrina Barry said the reason all four mission statements sound alike is because that every OTA is chasing the same underlying shift in customer behaviour
Packaging a concert into a flight, and selling out in 10 minutes
Turning to experiences, Currie asked Ru Yi to unpack what “collaborating with local partners to design integrated travel experiences” actually looks like in practice.
Her answer – securing additional seat capacity or flight frequency from airlines into a destination, coordinating with local tourism bodies so marketing efforts amplify each other rather than compete, and bundling hotels, flights and attractions into single packages promoted heavily on social media – on the logic that no single hotel or airline can make a destination trend on its own; it takes everyone pulling together.
The results she cited were striking: roughly 20,000 combined hotel-and-BTS-concert-ticket packages sold in Hong Kong within minutes, more than 6,000 similar packages moved in Thailand and Malaysia, and a Blackpink-linked package in Singapore the previous year that sold out its entire inventory in 10 minutes.
Karshagen’s parallel example centred on Expedia’s activities and experiences business – a segment he expects to keep growing, built on strong museum and cultural content, and one that increasingly feeds Expedia’s B2B strategy by giving partners a way to layer additional travel services on top of core bookings rather than just distribution alone.
On inbound-market potential for Queenstown, the panel’s picks scattered across India, Canada, Japan and South Korea, underscoring the global spread of the four OTAs on the panel.
