Booking.com is leveraging generative artificial intelligence to gain a deeper understanding of traveler intent, enabling more personalized content and improved user experiences. Speaking at ITB Berlin, Matthias Schmid, Senior Vice President of Accommodations, emphasized that AI is not just an evolution but a revolution, positioning the company at the forefront of the technology’s application in the travel sector.
The company has been utilizing machine learning for over a decade, with hundreds of models in place for enhancing offer relevance, language translation, pricing optimization, and fraud detection. It now categorizes generative AI applications into three key areas: improving internal productivity, refining machine learning models, and enhancing traveler engagement.
Internally, AI-powered coding assistants are helping software engineers, while automated tools summarize customer service calls, increasing efficiency. Booking.com is also encouraging employees to adopt generative AI to streamline workflows. Externally, the technology is making machine learning models more agile and improving how traveler intent is understood. This allows the company to personalize hotel descriptions based on user profiles, such as distinguishing between solo female travelers and families with children to provide more relevant property recommendations.
Schmid noted that while online travel agencies have traditionally excelled in search, booking, and payment processes, AI is now transforming the planning phase. Previously, users would turn to Booking.com only after deciding on a destination and travel dates. With generative AI, travelers can engage with the platform much earlier, receiving tailored recommendations and insights.
This approach aligns with strategies outlined by competitors such as Expedia Group, which has also been investing in AI-driven efficiencies and customer value enhancements. For Booking.com, AI offers significant benefits not only to users but also to partners. The company, which previously relied heavily on property descriptions provided by hotel partners, can now supplement this data with insights drawn from images, customer reviews, and other sources.
Since the launch of its AI-powered trip planner in June 2023, Booking.com has introduced smart filters and property Q&A features, allowing users to refine their searches and interact more effectively with listings. The functionality is currently available in the United States, United Kingdom, Australia, New Zealand, and Singapore, with plans for expansion across Europe. Additionally, the company is preparing to launch a trip assistant for post-booking support and a car rental pick-up assistant to help travelers navigate documentation requirements.
Schmid also discussed experiments with agentic AI and Booking.com’s collaboration with OpenAI. He described trials where an AI agent was tasked with completing reservations, allowing users to observe its decision-making process. While the technology still has some limitations, he expressed confidence that these challenges would be resolved within six months. Such experiments, he explained, provide valuable insights into potential friction points in the user journey, highlighting areas for improvement.
The evolving role of search in the AI landscape remains a critical focus for Booking.com, particularly given its investments in paid search and Google integrations. While the long-term impact of generative AI startups on search behavior is still uncertain, the company has integrated multiple large language models and built a dedicated AI integration layer over the past 18 months. This system ensures privacy protections, prevents hallucinations, and facilitates AI adoption across the organization in a controlled environment.
Schmid acknowledged the growing debate around data ownership, noting that companies must differentiate between proprietary data and publicly available information already being scraped by AI platforms. He emphasized that customer behavior, transaction history, and search patterns remain Booking.com’s most valuable data assets, making industry-wide collaboration in this area unlikely.
Reflecting on whether generative AI could level the playing field for smaller companies, Schmid pointed out that while certain applications, such as translation services, could benefit all players, more complex use cases require significant investment and expertise. He concluded by stressing that AI is no longer optional and urged the travel industry to embrace experimentation with the technology to stay competitive.
