# Post-Booking Servicing Is Where AI Will Deliver Its Highest Near-Term ROI in Travel
The Claim
While most AI discourse focuses on the glamorous end of travel — discovery, inspiration, booking — the real money may be in what happens after the confirmation email. This hypothesis argues that AI will generate its highest and most measurable returns in the 2025–2027 window not by reinventing the booking experience, but by automating the industry's most costly and stubbornly manual process: post-booking servicing. Changes, cancellations, disruptions, refunds, exchanges, and schedule modifications collectively cost the industry an estimated $200 billion per year in support operations.
The Evidence For
No hypothesis in the Phocuswright 2025 corpus is more consistently supported across multiple independent speakers. Riccardo Vittoria of Acai Travel — whose startup won the People's Choice Award — opened with the statistic that travel companies collectively spend $200 billion per year managing post-booking complexity, and reported live deployments achieving 60% handling time reductions with 20–30 basis-point CSAT improvements. NDC implementation has amplified the problem: handling times at call centers are up 20–30% due to NDC complexity, while experienced agents are retiring faster than replacements can be trained. Acai's three-agent architecture (supervisor, front-office, back-office) is designed precisely for this constraint.
Amir Balaish of Wenrix (runner-up Travel Innovator of the Year) added a quantitative dimension that crystallizes the inefficiency: 60% of in-flight servicing cases are 'complex edge cases' that consume 82% of total agent time. DeepFlow, trained on 50 billion real-world servicing data points, has achieved 93% automation for specific partner deployments — more than double the industry average. Integration time is one month.
Amadeus's Robert Buckman singled out travel disruption management as 'the next major frontier likely to be meaningfully solved within a few years' and made a rare move for a GDS executive: endorsing a specific startup (Acai Travel, a minority Amadeus investment) as evidence of the transformation underway. On Wall Street, Emma Taylor of Barclays identified call center automation as generating the 'most measurable near-term AI returns' in the entire sector — not booking reinvention, not personalization, but operational cost take-out with provable ROI.
Even executives whose focus lies elsewhere converged on this point. American Airlines' Neil Geurin explicitly identified post-booking, pre-travel as the most technically accessible personalization window because millisecond latency requirements don't apply. GBT's Evan Konwiser noted that within twelve months, corporate traveler expectations had shifted from demanding human agents in 30 seconds to accepting AI-first service pathways — a change in consumer psychology that removes the primary barrier to deployment.
The Evidence Against
The counterarguments are primarily about timing and data readiness rather than direction. Sabre's Kurt Ekert focused his conference narrative on agentic AI as a new distribution channel rather than servicing automation — suggesting strategic energy remains partially bifurcated between cost reduction and growth investment. Adam Harris of Cloudbeds noted that hotel data is still not in a shape that can power agentic servicing conversations at the property level, meaning small and independent hotels will lag behind major chains and OTAs in capturing this ROI.
Assessment
This is the most strongly supported hypothesis in the 2025 conference corpus. The convergence across speakers — startup pitches, incumbent executives, infrastructure providers, and capital markets analysts — is unusually high. The numbers are concrete and the deployments are live. The only substantive qualifier is a data readiness caveat at the property level, which delays but does not undermine the thesis. Airlines, OTAs, and TMCs face no such barrier: the data infrastructure exists and the ROI is demonstrable. Post-booking servicing AI is not a bet on the future — it is the clearest present-tense commercial opportunity in the sector.
**Verdict: Supported. Confidence: High.**