Travel After Search: AI and the Next Platform Shift

Travel has always occupied a unique place in consumer behavior because people rarely experience it as pure consumption. They experience it as transformation, identity, aspiration, and memory.

Goethe captured this sentiment during his Italian Journey, writing that he "would simply have perished" had he not made the trip. A quote which would feel relevant today as an instagram caption for a European summer post.

Travel remains one of the most emotionally protected categories of consumer spending. 63% of American travelers reported cutting back on everyday purchases including dining out and shopping in order to preserve their ability to travel. Even amid economic uncertainty, consumers continue to treat travel as non-negotiable.

With the demand side of travel remaining remarkably resilient, what is beginning to change with AI is the interface through which travel is discovered, planned and booked. More than half of Gen Z travelers now use short-form social video during trip planning, while millennials have become the leading adopters of AI tools for itinerary creation, hotel discovery, and destination research.

AI is fundamentally changing the architecture of a category expected to grow to $329 Billion by 2035 in the US alone. For a new wave of AI native founders, a window of opportunity is opening to disrupt the major players through owning, personalizing and automating the experience. Below are the four areas where I believe the next generation of breakout consumer AI companies in travel will emerge, but first let's go back to 2005.

Online travel has been one of the most successful categories in the history of consumer investing.

Looking back at the mid-2000s, it is remarkable how much of the modern internet economy was shaped by travel companies competing for distribution. Google became one of the world’s most powerful advertising businesses in part because travel companies became some of the internet’s most sophisticated performance marketers.

The year 2005 marked a major turning point. At the time, incumbents including Expedia and Travelocity were heavily focused on the “merchant model,” where OTAs packaged inventory, controlled pricing, and extracted high margins from suppliers. Booking.com took a very different approach.

As Bill Gurley pointed out in his 2013 A Rake Too Far essay, rather than maximizing take rates immediately, Booking embraced a lower-friction agency model that gave hotels greater control over pricing and inventory. Independent hotels could update availability directly through Booking’s extranet and immediately see how adjustments impacted search placement. The simplicity of the model proved extraordinarily powerful.

Booking rapidly expanded supply across Europe by onboarding long-tail independent hotels that larger OTAs struggled to serve efficiently. More supply improved consumer selection, which improved conversion, which justified increasingly aggressive search engine marketing spend.

This created one of the most important flywheels in internet marketplace history:

• broader inventory
• stronger conversion
• more efficient customer acquisition
• greater supplier adoption
• and ultimately dominant scale.

Booking.com understood something fundamental about online travel: distribution and conversion mattered more than controlling inventory directly. Its mastery of search marketing was frankly, legendary. Travel v1 became an extraordinary performance marketing business optimized around aggregation and intent capture.

But the architecture of travel is changing again.

TLDR: The next decade of travel may be defined less by who owns inventory and more by who owns orchestration.

For the past twenty years online travel was built around one core promise: comparison. Consumers opened dozens of tabs, compared flights and hotels across aggregators, scanned reviews, cross-checked maps, and manually stitched together itineraries from fragmented pieces of information. The dominant design of online travel startups became optimization through search.

The internet solved access to supply, but it never solved the cognitive complexity of travel itself. Flights fragmented into endless fare classes. Pricing became dynamic and opaque. Loyalty ecosystems multiplied. Consumers learned to navigate points arbitrage, hidden pricing logic, and constantly shifting inventory systems. Travel became easier to book and harder to optimize.

AI changes that architecture because it shifts the primary activity from comparison to orchestration. Instead of returning links, generative systems return plans. A traveler can now ask for a five-day Sicily itinerary optimized for minimal transfers and family hotels, or a Tokyo itinerary weighted toward food and walkability, and receive a coherent draft rather than a collection of search results.

This matters because the entity generating the most coherent plan increasingly controls the shortlist. When travelers receive a thoughtful itinerary with personalized aesthetic consistency, pricing considerations for accommodation and flights already embedded, alongside an ability to book, most users will just refine the draft rather than restart the search process. The itinerary itself becomes the interface as with companies like Odessia.

TLDR: The future travel interface may resemble a hybrid of TikTok, Pinterest, Google Maps/Flights, and ChatGPT more than a traditional OTA.

Travel platforms have historically relied on explicit filters such as price or star rating, but real travel decisions are shaped by emotion and taste.

Some travelers consistently choose boutique properties with strong design identities, even at a premium. Others prioritize predictable luxury brands. These are not isolated preferences. They are patterns. AI-native travel products can therefore begin to model:

• tolerance for layovers versus preference for direct flights,
• willingness to trade price for calm or convenience,
• aesthetic preference toward certain property types,
• and increasingly complex loyalty and rewards behavior.

This creates the possibility of persistent travel memory systems. Instead of treating each trip as an isolated search session, platforms can accumulate preference data over time, gradually building a much deeper understanding of how a traveler wants to move through the world.

One of the most interesting unlocks may come from identity and account integration. Rather than manually inputting preferences, travelers could log in with their email, airline, hotel, or credit card accounts, allowing systems to automatically infer patterns from past trips, loyalty behavior, receipts, upgrades, and booking history. In some cases, travelers may even delegate these credentials to AI agents, like Arden, that can monitor pricing, coordinate itineraries, or proactively optimize travel decisions on their behalf.

Over time, this creates the foundation for proactive travel planning rather than reactive search. An AI-native travel platform could recognize that a user has no plans over July 4th weekend, understand their historical preference for boutique beach hotels and direct flights, and proactively send a personalized itinerary or price alert over text or email before the traveler even begins searching.

The competitive advantage in travel may therefore shift from inventory aggregation toward understanding.

TLDR: The next generation of travel platforms will not compete solely on who has the most supply. They will compete on who best understands the traveler.

One of the most interesting signals in travel over the last several years has been the rise of companies like Fora. Fora did not invent travel advising, but it modernized it by building a digitally native infrastructure around independent advisors with software, community, and shared commissions.

In many ways, it formalized what used to be informal “mom travel agents” into scalable micro-entrepreneurs.

This development reveals something important about travel: consumers still value trusted curation, particularly for high-stakes, expensive, or emotionally important trips.

AI does not eliminate this layer. It amplifies it. Advisors equipped with generative systems can draft itineraries faster, optimize pricing more effectively, automate operational coordination, and serve more clients per head. The result is not disintermediation but leverage.

We are likely to see increasingly verticalized advisor ecosystems emerge around specific identities and constraints:

• families traveling with young children,
• multi-generational travel,
• wellness-oriented trips,
• luxury points optimization,
• remote work and team offsites.

In this model, AI handles orchestration and information processing while humans provide trust, taste, judgment, and emotional reassurance.

TLDR: The human travel agent returns not as a gatekeeper of inventory, but as a curator amplified by AI, or more likely not as a human at all, but as a fully autonomous agent.

One of the structural weaknesses of online travel v1 is that it optimized heavily around discovery while leaving pricing complexity largely unresolved.

Travel pricing has evolved into one of the most opaque consumer markets in existence. Fare classes shift constantly. Award availability changes dynamically. Loyalty systems create hidden pools of value. Consumers increasingly suspect that prices vary not only with supply and demand, but also with timing, behavior, and profiling.

The result is a system that appears transparent on the surface while remaining extraordinarily difficult to optimize beneath it.

This creates a major opportunity for AI-native optimization layers and reverse engineering.

Companies such as Roame and Miso are already building products focused on:

• points maximization,
• transfer arbitrage,
• reward routing,
• fare monitoring,
• and post-booking optimization.

The opportunity is not merely helping consumers discover where to go. It is ensuring that once the right option is identified, the traveler never overpays for it. Rebooking during disruptions, refund monitoring, cancellation tracking, and itinerary coordination remain largely manual.

TLDR: Travel is fundamentally a category filled with fragmented systems, hidden rules, and constant change. AI excels in exactly those environments.

AI changes the core unit of value in travel from inventory access to intelligent coordination. Discovery, pricing, loyalty optimization, and logistics increasingly collapse into a single, personalized and automated system.

However, foundational model companies such as OpenAI, Anthropic, and Gemini represent a very real long-term competitive threat to AI-native travel startups. At the same time, platform shifts consistently create temporary windows where new consumer behaviors form before incumbents fully consolidate them. There is likely a meaningful opportunity for startups to build both highly automated and highly visual, AI-native travel experiences that integrate discovery, planning, booking, pricing optimization, and post-booking coordination into a much more coherent consumer product.

The companies that ultimately win may not be the ones with the most inventory. They may be the companies that best understand how a person wants to move through the world and continuously optimize around that understanding.

Goethe travelled because he believed the journey would remake him, and he was right. The desire he described as a disease is now a mass-market expectation, protected in the household budget ahead of almost everything else, and it is about to be served by a completely new interface.

We think this is one of the most exciting consumer opportunities of the next ten years. If you are building it, please get in touch.