More than two decades ago, the rise of the internet economy produced a new generation of intermediaries that reshaped industries. Aggregators like Expedia and Booking.com in travel, Zillow in real estate, and Indeed in recruiting upended established channels, creating digital marketplaces that connected fragmented supply with mass demand. These aggregators built their power by monetizing search traffic through ads and transaction fees. They became gateways of the web, turning discovery itself into a business model that reshaped how people searched and made decisions online. Whoever controlled that discovery captured the revenue, which is why search has been the foundation of digital commerce. That model now faces its own test. Generative AI is changing how people get information online. Apps like OpenAI’s ChatGPT and Google’s Gemini engage users in conversation and generate direct answers rather than just lists of links. By reducing the need to click through to other sites, these systems are beginning to displace the gateway function that aggregators depend on, eroding their ability to capture intent at the top of the funnel, the foundation of their traffic and revenue. Early signs of disruption are already visible online. Education platform Chegg, which relies on search traffic to drive online tutoring, textbook rentals, and subscriptions, has reported sharp declines in visits since AI tools began answering students’ questions directly. The company has attributed the drop to users getting what they need inside AI interfaces rather than clicking through to its site, showing how quickly value can drain away when discovery moves outside the platform. AI Is Scrambling the Online Travel Model Online travel is a vivid test of how AI could disrupt the aggregator model. Expedia and Booking.com dominate digital travel, operating platforms that control about half of the industry’s global bookings. Like all aggregators, scale is crucial. The more travelers a platform attracts, the more suppliers it draws in; and the more suppliers participate, the more essential the platform becomes to travelers. That scale gives aggregators leverage with suppliers, allowing them to negotiate better terms, capturing more commissions from each booking and charging higher advertising rates on their sites. To sustain this scale, online travel agencies seek to capture demand at the earliest stage of discovery by paying for prominent placement in search engines like Google, which controls 90% of global search. For Expedia and Booking.com, this strategy has long shaped their business model and helps explain why marketing is their largest cost, with a significant amount going to search visibility and related advertising. Expedia spent more than half its revenue in 2024 on marketing, about $6.8 billion, while Booking.com spent roughly 30% of its annual revenue on similar marketing investments, more than $7 billion. AI is beginning to scramble these economics. Google is keeping travelers inside its AI-enhanced search environment longer and helping them curate trips before they leave the results page. Consumers are also turning to apps like ChatGPT for trip planning. These tools will still send users to online travel agencies to complete bookings, but discovery is drifting away from aggregator websites, reducing the opportunity to shape intent and influence purchase choices. As more planning moves to AI interfaces, online travel agencies will have fewer openings to cross-sell products, weakening the network effects that support their scale. It might get worse. While generative AI tools help travelers shape a plan, search engines and startups are experimenting with agentic AI to strengthen that experience. These new AI capabilities can continue working without prompting once they understand the traveler’s goal, drawing on connected systems to adjust an itinerary when conditions shift. As this capability improves, it becomes possible for an entire trip to be planned and booked inside the AI environment, with no need to click out at all. This shift makes it easier for new entrants to define the planning journey, creating a new battleground over who guides travelers from intent to decision. Sam Altman, CEO of OpenAI, has pointed to this potential disruption, describing chat interfaces like his company’s ChatGPT as a way to cut out intermediaries such as online travel agencies and drive margins dramatically down, reducing what he called the “taxes” that make travel more expensive. The same challenge will confront every industry built on aggregation, from housing to hiring. AI in Online Travel: The Double-Edged Sword of Disruption As AI reshapes the rules of customer acquisition and platform power, some of the original internet disruptors must navigate new threats and opportunities. Opportunities Threats AI can enable personalization at scale to drive more tailored recommendations and upsells. AI can enable new customer experiences with more relevant and contextual results that can increase conversion. Online travel agencies may be able to optimize for AI search in the same way they optimize for search engine traffic. AI lowers barriers to entry in all parts of the ecosystem. AI can enable suppliers to do more direct bookings themselves. If consumers flock to AI interfaces, top-of-funnel traffic could suffer and the booking process may become commoditized. Incumbents may struggle to manage investments in the front-end user experience while continuing to invest in core operations. Source: Innosight Disruption can take different forms, often emerging when new experiences redefine how and where consumers make choices. When incumbents rely on legacy business models that once made them strong, that success can become their weakness. The digital marketplaces that disrupted others now face the same challenge they once posed: how to adapt when the model that made them essential begins to move beyond them. In the online travel business, incumbents like Expedia and Booking.com will have to make three shifts to stay powerful in an AI-first market. 1. From Functional to Social and Emotional Jobs Once efficiency becomes automatic, advantage moves to what technology can’t easily reproduce. In the jobs-to-be-done framework—where customers “hire” a product or service to accomplish a goal—a functional job is the practical task they need completed, while emotional and social jobs describe the confidence and reassurance that influence which option feels right. For travelers, booking a trip is about more than finding the best price; it’s also about knowing recommendations reflect what they care about and having confidence the plan will work. Expedia and Booking.com grew by solving the functional jobs in travel that matter to customers. They amassed portfolios of brands that appeal across the market, serving different kinds of travelers and trip purposes. They also developed the infrastructure that makes online travel planning and booking feel safe and dependable: links to supplier systems, payment capabilities that work across markets, and service operations that step in when plans change. Their growth came from delivering reliability at moments when travelers needed it most, and that still shapes why so many people turn to them when they are ready to book. However, AI is changing the fundamental consumer experience. As generative and agentic tools make planning faster and more conversational, the early stages of the journey are shifting in a way that makes functional depth matter less and emotional resonance matter more. Travelers want plans that reflect their tastes. They want reassurance that a suggestion fits what they value. They look for experiences they can imagine themselves in before they commit. Those social and emotional jobs may now shape planning long before a booking ever hits an online travel agency. AI-native firms are building services around that distinction. Airial invites users to describe what they want, then generates a bookable itinerary within the same chat interface, compressing multiple steps into one seamless flow. To refine plans, it lets users link to social media videos of destinations and is piloting a revenue-sharing model with creators whose content drives users to its platform. Fora is taking a complementary path, equipping human advisors with AI tools that automate research and logistics so they can focus on tailoring experiences. Thousands of independent planners use its shared technology to deliver personalized guidance at lower cost. The incumbents are bringing more of that dimension into their websites and apps. Booking.com is expanding its Trip Planner, built on years of booking and review data, to anticipate traveler needs and refine suggestions. Expedia is embedding generative tools in its platforms, adding an AI agent to its Hotels.com app and incorporating features that enable users to upload social media videos of desired destinations. These moves show an intent to evolve, but the work is still limited in scope and geography, and the experience remains significantly less fluid than what newer entrants are already testing. Online aggregators built their dominance by making it easy to navigate vast travel options and trust that the final choice would work. But that formula is no longer enough. As AI reshapes how people plan and make decisions, the edge is shifting to platforms that deliver connection and inspiration. Incumbents must reinvent the experience or risk becoming irrelevant in a market that will reward emotional engagement. That means rethinking the user journey from the ground up, infusing discovery with personalization and a sense of narrative, rather than just optimizing search by layering AI onto an old model. 2. To an AI-Driven Ecosystem AI is rewriting how travel is planned, but incumbents still start from a position of tremendous strength. Expedia and Booking.com are a critical part of the travel ecosystem because they provide the booking and operational backbone that other services need. Their systems link into airlines and hotels, handle payments across markets, and manage the work required to keep itineraries intact when plans change. That infrastructure is what lets search engines and emerging AI assistants turn a user’s interest into an actual trip—which is why these platforms continue to rely on incumbents even as they experiment with new discovery experiences. That strength explains why Google is integrating with the incumbents even as it pushes into agentic trip planning. The company recently announced that its AI Mode will help travelers build itineraries and complete bookings directly inside its Search function through a new conversational workspace. Google said it is partnering with Expedia and Booking.com so these bookings can be completed through their systems. Even ChatGPT depends on the same foundation; its travel features surface options inside the chat interface, but it recently formalized its relationship with aggregators to provide real-time pricing and a seamless booking experience. Expedia and Booking.com are leaning into this strategy, carving out a role as gateways within these conversational environments. In addition to Google and ChatGPT, other AI-powered search engines such as Perplexity have announced partnerships with the aggregators, positioning them as the safe, reliable execution layer behind generative and agentic AI. As discovery shifts into external chat interfaces, the incumbents are moving from buying placement in search to negotiating their presence inside these AI environments. For instance, Google has said the arrangement inside AI Mode follows the same setup, where the online travel agencies pay for visibility and earn commissions when a booking is completed. Executives at the incumbents acknowledge that this shift demands a strategy of aligning with the platforms shaping the new interface. Booking Holdings CEO Glenn Fogel recently explained: “Our belief is that as long as we work closely with them, we will be able to participate in a way that provides a great return for our customers and our partners.” Yet this strategy exposes a deeper tension: The economics that made these companies powerful are now at risk. The aggregator model has depended on reaching travelers early in their journey and converting engagement into volume and margin. But as discovery moves onto third-party platforms, incumbents may see shrinking revenue per customer, even if overall volume holds. They are also betting that today’s dominant platforms will continue to anchor the search experience, even as AI reshapes how travelers discover and plan trips. Expedia and Booking.com are experimenting with ways to improve the user experience on their own websites, but those efforts now run parallel to a shift in where intent is formed. The uncertainty does not end there. New agentic tools are being designed to work across devices and connect directly to supply, giving hotels and airlines a potential channel that bypasses online travel agencies entirely. Suppliers are watching these developments closely and testing early versions of these systems. The incumbents still provide the infrastructure that powers online travel, but that position will be harder to defend as new interfaces gain autonomy. To stay essential in this ecosystem, online travel incumbents will need to evolve beyond fulfillment and shape an experience that travelers and suppliers actively choose, because the advantage in an AI-driven market will come from the strength of that relationship. 3. From Structured Data to Interpreting Intent For years, online travel agencies gained an edge by owning large datasets and optimizing transactions. Every booking, review, cancellation, and service interaction created the foundation for the recommendations travelers see today. This depth let Expedia and Booking.com organize vast amounts of supply, keep prices accurate, and deliver reliable choices. AI is beginning to reshape how that advantage works. Large language models can interpret traveler intent without depending on any single platform’s structured data. New interfaces can blend information from suppliers, review sites, maps, and user content in one view. Discovery no longer favors the biggest dataset; it depends on who can turn ambiguous intent into a set of choices that feel right. That moves the center of gravity from aggregation to curation, shifting power toward systems that can make sense of signals in real time. AI-native travel firms such as Airial and Fora are moving in this direction. They capture customer intent through direct interaction, gathering signals that never appear in traditional clickstreams: what travelers hope a trip will feel like, what inspires them, what tradeoffs they’re willing to make. Airial uses prompts and social content to close the gap between inspiration and booking. Fora turns conversations with advisors into guidance shaped by personal history, not generalized patterns. Their data advantage is proximity to the traveler’s motives, not the size of their inventory. The incumbents are bolstering their position by unifying data so AI systems can generate clearer, more timely recommendations. Expedia is consolidating siloed data across its brands so its AI can act on live booking, payment, and service signals. Booking Holdings, the parent company of Booking.com, Priceline, Kayak, Agoda, and OpenTable, is making similar changes, creating shared access to behavior and performance data across its portfolio so that its AI Trip Planner can refine suggestions with more context. These efforts aim to turn large, previously disconnected datasets into engines for sharper recommendations and faster problem-solving, making curation a competitive strength, rather than a weakness. Still, challenges persist for the incumbents. They hold the industry’s deepest operational datasets, yet curation depends on using that information to offer clear guidance at the exact moment intent forms. Their data work gives them the foundation to do this, but the real test is whether they can translate operational depth into interactions that keep travelers engaged before intent shifts to an external AI interface. In a market where AI increasingly interprets demand, advantage will go to those who show up at the start of the journey and keep travelers engaged all the way through to the booking. . . . AI is fundamentally reshaping the online travel ecosystem by shifting where value is created and captured. The real disruption lies in how search and discovery is being transformed, as AI empowers consumers in new ways, resets expectations, and enables new entrants to create novel value propositions. The critical battleground is the consumer experience layer. The online travel incumbents, whose competitive moat was built on maximizing their websites’ visibility in a search engine’s results through massive marketing spend, are taking logical steps to defend the current model. It’s possible they and partners like Google adapt fast enough to stay in control of the online travel ecosystem. But value will quickly migrate away from transactional efficiency to personalized and nuanced experiences. Whoever gets there the fastest will win.