By Global Travel Technology Correspondent
Published: October 2026
Executive Overview
The global online travel agency (OTA) landscape is undergoing its most profound structural transformation since the advent of mobile booking applications a decade ago. At the heart of this disruption is generative artificial intelligence (GenAI) and its evolution into autonomous agentic systems. Two industry titans—Booking Holdings and Expedia Group—now find themselves locked in a high-stakes technological arms race. Both conglomerates possess the balance sheets, proprietary data lakes, and massive consumer reach required to approach consumer-facing AI from multiple vectors simultaneously. Their strategies encompass conversational chatbots, deeply specialized task-oriented agents, and internal venture-style incubation hubs.
While their overarching goals overlap—principally, capturing downstream intent, streamlining customer friction, and driving conversion rates—their strategic execution reveals distinct corporate philosophies. Booking Holdings, under the leadership of CEO Glenn Fogel and CFO Ewout Steenbergen, is pursuing a decentralized, multi-pronged approach involving at least five distinct AI initiatives. This includes incubating internal startups, leveraging brand-specific implementations like Priceline’s revamped "Penny" agent, and preparing a suite of upcoming deployment tools across powerhouses such as Booking.com and Agoda.
Conversely, Expedia Group has focused heavily on creating unified underlying tech stacks, leaning into its OneKey loyalty ecosystem to synthesize consumer data and power cross-platform AI assistants. As these systems graduate from experimental novelties to core revenue drivers, the battle for the AI-enabled traveler is no longer a distant theoretical exercise. It is happening in real time, reshaping customer satisfaction scores (CSAT), operational cost structures, and the very definition of online trip planning.
Detailed Chronology: The Evolution of OTA Artificial Intelligence
To understand where Booking Holdings and Expedia stand today, one must examine the rapid acceleration of AI deployment across the travel sector over the past thirty-six months.
Phase 1: The Rule-Based and Early LLM Era (2022–2023)
In the immediate aftermath of OpenAI’s public release of ChatGPT in late 2022, the online travel sector experienced an existential awakening. Legacy chatbots—rigid, rule-based systems that repeatedly forced frustrated consumers into unhelpful customer service loops—were rendered obsolete overnight. Both Booking and Expedia rushed to announce partnerships with foundational model providers.
During this initial phase, the primary use case was experimental. OTAs integrated basic large language model (LLM) wrappers to answer simple FAQ queries, suggest destination highlights, and assist with rudimentary itinerary generation. However, these early deployments suffered from well-documented limitations: "hallucinations" regarding hotel availability, high latency, and an inability to execute transactional tasks (such as modifying a flight or processing a refund) without handing the user off to a human support agent.
Phase 2: The Shift Toward Agentic Workflows (2024–2025)
Recognizing that mere chat interfaces were insufficient for driving bottom-line metrics, engineering teams at both Booking Holdings and Expedia pivoted toward "agentic" AI architectures. Rather than acting strictly as conversational partners, agentic systems are designed to reason, plan, and execute multi-step workflows across disparate enterprise databases.
During this period, Booking Holdings restructured its internal product pipelines. Instead of relying solely on a monolithic corporate AI strategy, leadership encouraged individual brands—Priceline, Agoda, and Booking.com—to build tailored solutions. Priceline took an early lead by completely re-architecting its conversational assistant, Penny, transitioning it from a static FAQ bot into an underlying agentic engine capable of contextual search, dynamic price negotiation simulations, and personalized recommendation filtering.
Phase 3: Commercial Scale and Incubation (2026 and Beyond)
By 2026, the technology matured past the proof-of-concept stage, entering an era of rigorous commercial accountability. Speaking at recent investor and financial conferences hosted by major financial institutions, executive leadership from Booking Holdings outlined a remarkably aggressive, multi-layered portfolio strategy. Rather than placing all their bets on a single consumer-facing interface, Booking Holdings revealed that it is actively advancing five core AI initiatives simultaneously.
This multi-pronged approach includes establishing two internal startup incubators tasked with testing radical user interfaces outside the constraints of legacy website frameworks. At the same time, core brands are rolling out production-ready features. Agoda, known for its dominant footprint in the Asia-Pacific (APAC) region, has developed a proprietary trip-planning tool slated for commercial launch in the near term, building upon localized testing phases that commenced late last year.
Supporting Context & Metrics: Evaluating the Impact on Conversion and CSAT
In the cutthroat world of digital travel distribution, technological innovations are ultimately judged by hard financial metrics: conversion rates, average order value (AOV), customer acquisition costs (CAC), and customer satisfaction scores (CSAT).
The Metrics of Success
While OTA executives are notoriously protective of granular proprietary data, recent disclosures provide a compelling window into the early economic returns of agentic AI deployments.
According to statements delivered by Booking Holdings CEO Glenn Fogel at high-profile financial symposia, preliminary performance indicators for the revamped Priceline Penny agent are notably robust.
- Customer Satisfaction (CSAT): Fogel noted that the company is "definitely getting good CSAT numbers" from users engaging with the agentic iteration of Penny. In an industry where customer service interactions are historically a source of high friction and consumer churn, measurable improvements in post-interaction sentiment represent a significant competitive moat.
- Engagement and Conversion: Beyond qualitative satisfaction, early conversion data reveals an encouraging correlation. Fogel pointed to "good engagement and higher conversion" rates among users who interact with the AI assistant during the inspiration and booking funnel.
The Caveat of Scale
Despite these promising green shoots, corporate leadership remains appropriately cautious regarding extrapolation. Fogel was careful to contextualize these early wins, explicitly reminding financial analysts and institutional investors that these metrics derive from a "very small sample" of total platform traffic.
Scaling these early successes across Booking Holdings’ vast global ecosystem—which processes hundreds of millions of annual active users across Booking.com, Priceline, Agoda, Kayak, and OpenTable—presents massive infrastructural and computational challenges. Ensuring that low-latency responses, transactional security, and hyper-personalized inventory matching remain consistent at scale requires continuous cloud optimization and billions of vector database queries.
Official Statements: Insights from the C-Suite
The strategic divergence and shared ambitions of Booking Holdings’ executive team were laid bare during a flurry of presentations at recent global technology, media, and telecommunications (TMT) investor conferences.
Glenn Fogel on the Power of Experimentation
Addressing attendees at the Citi Global TMT Conference, Booking Holdings CEO Glenn Fogel emphasized that the company’s organizational structure is specifically designed to prevent bureaucratic stagnation in the face of rapid technological evolution.
"Both Booking and Expedia have the luxury to address consumer-facing AI from multiple angles, including chatbots, specialized agents, and startups. Their approaches overlap, but also have distinct differences," Fogel noted, framing the competitive dynamic between the world’s two largest OTA giants.
Elaborating on Priceline’s traction with Penny, Fogel highlighted the strategic decision to move beyond generic text generation:
"We’re definitely getting good CSAT numbers. We’re in good engagement and higher conversion… [though] it’s a very small sample."
Ewout Steenbergen on Regional Deployments and Agoda’s Roadmap
Meanwhile, speaking at a concurrent investor conference hosted by Goldman Sachs, Booking Holdings Chief Financial Officer Ewout Steenbergen provided deeper color regarding the operational cadence of brand-level rollouts. Steenbergen confirmed that the conglomerate’s five-pillar AI strategy is yielding staggered product releases tailored to specific regional markets and consumer behaviors.
Highlighting the upcoming commercial deployment within the Agoda ecosystem, Steenbergen stated:
"Agoda will launch something similar in the near future."
This rollout follows a series of iterative backend tests where Agoda deployed specialized machine learning modules to optimize room inventory matching and localized payment routing across complex Asian travel corridors. By granting individual brand leadership teams the autonomy to experiment with distinct technical frameworks—backed by shared centralized AI infrastructure—Booking Holdings aims to avoid the pitfalls of a top-down, one-size-fits-all product strategy.
Future Outlook: The Next Frontier in AI Travel Tech
As the travel industry looks toward the remainder of the decade, several critical vectors will determine whether Booking Holdings, Expedia, or emerging pure-play generative AI startups capture the ultimate high ground of consumer travel intent.
1. From Search to Execution: The Autonomous Trip Architect
The holy grail of travel AI is the transition from a passive search assistant to an active, autonomous execution engine. Future iterations of tools like Priceline’s Penny and Agoda’s upcoming planner will not merely suggest a five-day itinerary to Tokyo or a boutique hotel in Paris. They will automatically reserve flights matching optimal carbon-offset profiles, secure dining reservations at fully booked restaurants through integrated API partnerships, negotiate dynamic rate discounts with properties, and autonomously manage real-time itinerary disruptions (such as weather-induced flight cancellations) without requiring human intervention.
2. The Consolidation of Loyalty and Personalization
Both Booking and Expedia understand that AI is only as powerful as the proprietary data feeding its algorithms. Expedia’s aggressive push around its unified OneKey loyalty program provides it with an integrated data flywheel spanning flights, hotels, and car rentals. In contrast, Booking Holdings leverages its unrivaled accommodation supply density. The future winner of the AI travel wars will likely be determined by which platform can most seamlessly synthesize historical consumer preferences, real-time contextual data (weather, local events, flight statuses), and financial incentives into a single, frictionless conversational thread.
3. Regulatory and Economic Pressures
As conversational AI agents assume transactional authority over billions of dollars in consumer spending, regulatory scrutiny regarding algorithmic bias, data privacy, and price transparency will inevitably intensify. Furthermore, the high computational costs associated with running massive agentic LLM workflows will place a premium on operational efficiency. OTAs that successfully balance aggressive innovation with sustainable unit economics will cement their market dominance.
Conclusion
The AI arms race between Booking Holdings and Expedia is far more than a battle of marketing slogans or superficial chatbot implementations. It is a fundamental rewiring of how humanity discovers, plans, and experiences the world. With initiatives like Priceline’s Penny demonstrating early success in conversion and customer satisfaction, and upcoming rollouts from Agoda poised to expand the frontier, Booking Holdings has positioned itself at the bleeding edge of travel technology. Whether this multi-pronged, startup-incubated approach will decisively outpace Expedia’s unified ecosystem strategy remains the defining question for the global travel sector in the years ahead.
