Travel & Tourism

“How was last month?” has several answers in a hotel. The property management system knows the stays, the channel manager knows bookings by channel, the portal has its own commissions, the booking engine on the website knows the visits before, and the review arrives weeks later.

We build a shared data foundation with consistent definitions for occupancy, price, revenue and channel. Business intelligence is at the centre, for hotels as much as for tour operators and destinations, only the metrics differ. Forecasts and AI assistants come on top, without decisions leaving the team’s hands.

Which booking counts when?

Does a booking count at the time it is made, at arrival or at payment? Is the price gross or net, with or without breakfast? That has to be settled before numbers are compared. For RevPAR, room revenue divided by available rooms, it matters among other things whether out-of-order rooms count as available and whether taxes and commissions are in the revenue. We set the definition once and calculate every report from it.

Then the metrics can be read over time. Booking lead time shows how early guests book, and pick-up shows how the on-the-books position develops against last year at the same cut-off date. For cancellations, look at channel, rate and period, because the average hides where the problem is.

What a channel really brings in

A channel with high revenue can contribute less than a smaller one after commission, cancellations and effort. You only see that when revenue, costs and cancellations per channel sit in the same model. That includes the direct booking share, meaning how many bookings come through your own website.

Booking.com has been designated a gatekeeper under the Digital Markets Act since 2024. As a result, the portal may not prohibit hotels in the EEA from offering rooms cheaper on other channels, direct sales included. Price comparison between channels becomes more important because of that. For competitor prices, use only publicly available prices, and agree the data sources with legal counsel beforehand.

The booking engine on someone else’s domain

When the booking engine runs on a vendor’s domain, the origin and consent of a visit are easily lost. Channel costs then stay an estimate, because you cannot say which page led to the booking. Server-side tracking can help if a shared booking ID links visit and booking. It also shows where interested guests drop out of the booking funnel.

Forecasts and price suggestions

A demand forecast uses on-the-books position, seasonal patterns and external factors such as school holidays, public holidays and weather. It helps with staffing and purchasing, and with pricing too. Trade fairs, concerts or roadworks are only known to it if they are in the data. We test forecasts on past periods and show the deviation openly.

For pricing, we recommend suggestions that revenue management approves. Automatic adjustments without limits lead to prices that do not fit the brand or the property. Floors, ceilings and special cases can be stored as fixed rules, and whoever approves the prices knows about groups and renovations.

Reviews and guest inquiries

Language models assign reviews to topics such as cleanliness, breakfast or location, and spot checks show whether the assignment is right. This makes visible which topics keep coming back and how they develop over time and across properties.

An assistant for guest inquiries answers questions about arrival, amenities or cancellation terms from your own content and names the source. Complex or sensitive cases go to staff. Since August 2026, the transparency duty under Article 50 of the EU AI Act applies to such chat assistants: guests must be able to tell that they are talking to an AI. Price recommendations are not among the high-risk applications in Annex III, but the classification depends on the actual use.

Guest data, registration duties and holiday rentals

Name, contact, payment and travel data are personal. For pricing and planning, aggregated data is usually enough. Consent and purpose limitation matter most for tracking and for profiling individual guests.

In Germany, the registration form (Meldeschein) has no longer been required since 1 January 2025 for guests with German citizenship. It still applies to foreign guests. Municipalities that levy a visitor tax need other ways to collect the data. Regulation (EU) 2024/1028 on short-term rentals has applied since May 2026. It governs registration and the transmission of data by platforms to authorities, and it mainly affects holiday rentals, not traditional hotels.