AI is changing travel discovery. Now it is changing travel commerce.

The travel industry has spent two decades optimising for search visibility. It may spend the next decade optimising for recommendation visibility instead.

The shift, from being found through search to being selected by AI, is already changing how travellers discover travel. We explored what it means for OTAs and tour operators in a previous article. Travelport's CTO Andrew Jordan framed it well, AI is already changing discovery, but booking remains the challenge.

But the supply side faces a different and arguably more commercially significant question. If AI increasingly decides which brands travellers consider before they ever reach a travel website, how do airlines, hotels, OTAs and technology providers capture the commercial value once those travellers arrive?

The advertising inventory, the data architecture, the distribution relationships — all of it was built for a world where travellers arrived via predictable channels. AI is changing where they arrive from, how qualified they are when they get there, and what they expect to find. The commercial infrastructure must follow.

OTAs: the media business becomes more valuable

AI threatens the OTA discovery model, that much is clear. When a traveller forms their shortlist inside an AI assistant before they ever reach an OTA homepage, the OTA loses the browsing phase its commercial model has traditionally depended on. But the same shift that threatens the discovery model simultaneously increases the value of the media business.

The travellers arriving at an OTA via an AI recommendation are not browsing. They have already narrowed their options, they are there to confirm and convert. The change in audience quality, from explorers to decision-ready visitors, makes OTA media inventory significantly more valuable to advertising partners.

A tourism board, hotel group or ancillary partner reaching a traveller who has already committed to a trip represents a fundamentally different commercial proposition from reaching someone still comparing options.

Most OTAs are not yet packaging their media inventory to reflect that difference. The AI discovery era is creating a media business worth considerably more than the one that preceded it but only for the businesses that build the infrastructure to capture it.

The opportunity isn't simply receiving better audiences, it's having the infrastructure to recognise their value and commercialise them accordingly. That's exactly the challenge Uplift was built to solve.

"OTAs have always sat on a valuable audience. What changes in the AI era is the quality of that audience at the moment they arrive. The ones coming via AI recommendation have already decided, they are there to book, not browse. That is an advertiser's dream but a lot of OTAs are not yet packaging it that way," says Roy Stratford, Director of Retail Media at Platform 195.

Hotels: what AI actually rewards

Where OTAs face a monetisation challenge, hotels face a visibility challenge. For years, hotel visibility has been driven largely by review volume and search optimisation. AI is beginning to reward something different: structured, consistent descriptions that clearly explain what a property actually is and who it is for.

Lighthouse's research, drawing on over 4,500 ChatGPT travel prompts across multiple destinations, makes this concrete. AI recommendations are highly concentrated — in Tokyo, only 10% of hotels were ever mentioned; in Paris, 13%. The majority never appeared at all but not because they were poor properties, rather the AI systems could not find sufficient, well-structured information to recommend them with confidence.

Lighthouse also found that review scores mattered less than structured, consistent property information. Star ratings and descriptive content proved stronger signals for AI recommendation than review volume alone.

The opportunity that unlocks for hotels that get this right is significant. Amadeus's early AI commerce pilots show that AI-driven visitors reached booking engines at 44.7%, compared with 25.9% from organic search. Amadeus's wider AI strategy, building an orchestration layer connecting hotel inventory directly to conversational AI flows, is a direct commercial response to that dynamic.

For hotel groups, the practical priorities follow from this: how properties are described across every digital surface an AI system might draw on; the data infrastructure that personalises the experience when that recommendation-led visitor arrives; and advertising inventory positioned close to the booking decision rather than deployed broadly across generic channels.

The opportunity increasingly lies in treating those environments as commercial assets rather than simply as booking channels. That requires the ability to package, manage and measure media inventory systematically rather than through fragmented partnerships.

Hotels have spent years optimising to be chosen by travellers. Increasingly, they also need to optimise to be chosen by the systems making recommendations on travellers' behalf. Those are not quite the same thing, and the hotel groups that recognise the distinction earliest are likely to benefit most as AI recommendation becomes a larger source of demand.

Airlines: owned environments become more valuable, not less

As more of the discovery journey moves into AI interfaces, fewer travellers will arrive at an airline's owned digital environment accidentally. Instead, they will increasingly arrive having already been guided towards a specific choice and with a clearer booking intent.

AI increases the commercial value of owned digital environments precisely because fewer travellers will arrive there accidentally.

That sounds counterintuitive, but it changes the economics of airline retail media. The question is not how to monetise a broad audience of varying intent, it is how to capture the maximum commercial value from a smaller, more qualified audience of travellers who are already close to a decision. Native, contextual placements that sit close to the moment of booking intent are worth considerably more in that environment than generic display served to a less certain audience.

Breaking Travel News's coverage of Phocuswright Europe identified AI-driven distribution and digital identity as two of the defining themes of the conference. For airlines they are the same problem because capturing the value of a pre-qualified traveller depends on the ability to recognise them.

First-party data, consented audience signals, and clean identity matching across markets are the foundations that make airline media inventory genuinely targetable rather than broadly served.

GDS and technology providers: the data layer is the distribution layer

For GDS businesses and travel technology providers, the AI discovery shift raises a question that goes to the heart of their commercial position.

If discovery increasingly happens inside AI interfaces that draw on structured, machine-readable data, then control of the data layer becomes a strategic asset in a way it was not before.

The PhocusWire whitepaper on the decade ahead for travel professionals identifies the agentic era as the next significant shift: AI agents connecting to live inventory, making recommendations, and facilitating transactions with increasing autonomy. The technology providers that position their infrastructure as the trusted, well-structured data layer that AI systems can reliably draw on are likely to become increasingly influential within a distribution ecosystem that relies less on traditional search and more on structured, machine-readable data.

In practice, that means structured inventory, clean APIs, machine-readable metadata and consistent data presentation across every interface an AI agent might query. The inventory AI systems can understand and verify is the inventory they surface. As the largest AI platforms increasingly favour structured, reliable data sources, providers treating data quality as a commercial priority, not simply a technical one, will be best positioned for this shift.

"The GDS model was built on being the most complete and reliable source of travel inventory. That value proposition does not disappear in an AI world, it becomes more important. But the format and structure of that data needs to be rebuilt for the systems that are now consuming it," says Roy.

The question the industry needs to answer

Hotels, OTAs, airlines, and GDS providers face different versions of the same underlying change. AI is delivering recommendation-led travellers to their commercial environments, visitors who have already been directed there, already formed a view, and arrive with a clearer purpose than most of the traffic that preceded them.

AI is not removing commercial opportunity from travel. It is concentrating it around businesses with the capability to recognise, understand and commercialise recommendation-led demand.

That infrastructure question, how to build, manage, and monetise media inventory that performs at the moment of intent, is what Uplift is designed to answer. Not as the conclusion to four separate arguments, but as the common response to a single commercial reality that is forming across the industry right now.

‍

This is the second article in Platform 195's series on AI and travel discovery. Read the companion piece, AI is changing who gets considered before a booking ever happens, which covers what the discovery shift means specifically for OTAs and tour operators.

To discuss how AI discovery is affecting your commercial position, get in touch with the Platform 195 team.

‍