What Can Travel Learn from AI Native Startups? #Phocuswright research
Summary
Mike Coletta, Senior Manager of Research and Innovation at Phocuswright, presents findings from a global survey of travel entrepreneurs conducted in May and June 2025, exploring how AI is reshaping the travel startup landscape at a time when venture funding has slumped to decade lows. The research reveals that 81% of travel startups are already using generative AI in some meaningful form, with only 5% having no plans to adopt it. The top internal use case is productivity (nearly two-thirds of respondents), followed by content creation (just over half), integration into core products for customers (exactly half), and customer support (around one-third).
The impact is overwhelmingly positive: 54% report major positive impact and 32% report minor positive impact, totaling 86% seeing net gains. The most notable benefit, cited by a full third of respondents as the number one positive impact, is faster product development and prototyping. Coletta argues this is inverting the classic lean startup model — whereas the prior wisdom was to validate market fit before building, AI now makes prototyping so cheap and fast that building first and testing resonance may be the superior approach.
Investment allocation data shows 58% of startups direct the majority of funds to product development and engineering, though Coletta flags this share may shrink as AI takes on more sophisticated technical tasks. A third of respondents expect at least 26% of their codebase to be AI-written or co-written within 12–24 months, echoing real-world benchmarks from Robinhood and Salesforce where AI already generates over 50% of new code. On competitive dynamics, startups overwhelmingly believe smart use of GenAI will be a major differentiator (ranked number one), and acceleration of product and feature launches was the top-ranked near-term impact (selected by nearly 70% in their top three).
Roughly a third believe AI-native startups could out-compete slower incumbents, while very few believe incumbents can entrench advantages simply by adopting the same tools. On the question of autonomous AI agents and agent-to-agent commerce, nearly all respondents consider these developments very relevant or game-changing for their business and the broader industry over a three-to-five-year horizon. Coletta closes with a speculative BCG-inspired scenario: an AI-only OTA launches with no human agents, promises 10-second refunds, zero fees, automatic disruption management, 1:1 loyalty optimization, and charges only a 5% commission — causing analyst downgrades to incumbent OTAs by noon.
He describes this as a possible but not certain endgame, noting the most likely OTA disruptor could be OpenAI rather than a startup. An audience poll showed a majority of attendees consider such a scenario at least possible within 10 years. Coletta's three recommendations for incumbent travel companies: become AI-adaptive by architecting highest-impact workflows around AI (not just adding it); double down on uniquely human capabilities like empathy and imagination; and position as a valuable node in the AI ecosystem by focusing on unique data assets, clean data, interoperability, and trust.
BCG estimates AI-only rivals are 5–15 years away, a timeframe Coletta notes has historically been sufficient for major industry disruptions in travel.
Key Takeaways
- •81% of travel startups are already using GenAI in some form, with 86% reporting a net positive impact — adoption is no longer early-adopter territory.
- •The fastest-growing benefit of AI for startups is speed of product development and prototyping, suggesting travel companies should prioritize AI integration in their build cycles.
- •The lean startup model may be inverting: AI makes prototyping so cheap that building a prototype first, then testing market resonance, can now outperform the traditional validate-first approach.
- •A third of travel startups expect AI to write at least 26% of their codebase within two years — plan now for what engineering teams look like in that environment.
- •Most travel startups believe they can out-compete incumbents via AI, and very few believe incumbents can entrench their advantage by adopting the same tools — structural speed and lack of legacy are genuine moats.
- •Nearly all surveyed founders see autonomous AI agents and agent-to-agent commerce as very relevant or game-changing for travel within 3–5 years — build interoperability and data readiness now.
- •The most likely major OTA disruptor may be OpenAI or a similar AI-platform company, not a traditional startup — incumbents should monitor platform-layer threats, not just startup competition.
- •Become AI-adaptive by architecting your highest-impact workflows around AI, not by bolting AI onto existing processes.
- •Strengthen uniquely human capabilities — empathy, imagination, and human touch — as a deliberate strategic counterweight to automation.
- •Focus on clean, interoperable, trusted data assets: being a valuable node in the emerging AI ecosystem is a defensible position for travel companies that cannot out-code AI-native competitors.
Transcript
Typically in our startup research, we focus on funding partly because it's the easiest yard stick to measure the landscape. [clears throat] But this year, as funding is faltering and as AI reshapes the industry, we want to dig deeper into what startups are doing and thinking. So we conducted a global survey of travel entrepreneurs. So today I'll be translating some of the results into takeaways for the industry. Survey was conducted in May and June of this year. Given how fast things are moving,...
