Based on the keynote by Cindy Heo, Professeur Associé, Ecole Hôtelière de Lausanne — Global Revenue Forum 2026.

In hospitality, the future of data analysis does not depend on the volume of information collected, rather on the ability to deeply understand its meaning.

Big Data tells us what guests do. Deep Data explains why, connecting behaviours, travel motivations and the context of the stay. This distinction is crucial for hotels, where needs vary depending on the purpose of the trip and the composition of the group. With AI, these data can be cross-referenced to anticipate needs, personalize the guest experience and identify new revenue opportunities.

In the hotel industry, AI is now part of everyday operations, even though its use is still often focused on efficiency and automation of administrative tasks. The real challenge, however, is to transform AI into a value-creation tool capable of innovating the Guest Experience and unlocking new revenue opportunities. To achieve this, hotels will need to move from generic Big Data to more specific and contextualized Deep Data, capable of turning guest knowledge into real value.

In the hotel industry, Generative AI is primarily used to automate administrative and operational tasks. An approach that optimizes efficiency, but not enough to generate incremental profits or create a true competitive advantage. The main limitations lie in two factors:

• Inability to create exclusivity: generative models process publicly available data (such as reviews on Booking.com or TripAdvisor), providing insights that are accessible to every competitor.
• Mathematical unreliability: generative AI still has significant limitations when it comes to quantitative elaboration. For this reason, professor Heo advises against relying on it for critical Revenue Management and pricing decisions, which require advanced analytical models based on the hotel’s proprietary data.

For decades, marketing and operations have classified guests through static demographic personas: the business traveler, the couple, the family.
In the AI era, however, this model reveals all its limitations:

• Personas are based on aggregated averages, while AI can identify micro pattern and behavioural signals.
• Guests are not static; their needs and willingness to spend change depending on the context. For example, the same traveler may seek value for a business trip and a premium experience for a family trip.
• Rather than simply knowing who the guest is, it becomes essential to understand why they are booking at that specific moment and under those specific conditions.

The new approach to segmentation, therefore, no longer starts with the guest’s identity, but with context and intent.

The gap between price and context becomes particularly evident at the front desk, especially when it comes to room upselling. Simply offering an upgrade at a fixed rate is often not enough. Without a perceived value that is aligned with the guest’s actual needs, the offer risks being rejected.
This is where Deep Data comes into play.
By combining information such as arrival time, purpose of the trip, booking lead time and the marginal costs of services, AI can estimate the likelihood of acceptance and suggest the most suitable offer for each guest.
An upgrade can therefore become more than a simple discount: a restaurant credit, an included service, flexible check-out, or a personalized experience can increase the perceived value while keeping the hotel’s marginal costs low.
For unsold rooms, which, once the night has passed, represent revenue that is permanently lost, this means turning perishable inventory into high-margin revenue, while simultaneously delivering a more relevant and personalized guest experience.

To build a future-ready operational AI model, hospitality leaders should align their operations to a Deep Data framework:

• Descriptive: identifies the real motivation behind guest behaviour, performance trends and churn, rather than simply focusing on transactional metrics.
• Experiential: identifies micro-preferences, behavioural signals and emotional triggers throughout the entire guest journey, from a 360-degree perspective.
• Exclusive: relies on proprietary internal datasets specific to each property, rather than public or generic industry benchmarks that can be easily replicated by competitors.
• Predictable: shifts hotel operations from reactive problem-solving to proactive, automated service design that drives long term revenue growth.

Transactional Big Data only records whether a guest said yes or no.
Deep Data reveals why they made that decision and how operations can influence it. The future of hospitality excellence lies in connecting guests’ contextual signals to create hyper-personalised experiences that drive satisfaction, operational efficiency and long term profit growth.