This insight summarises What is Ontology for Hotels by Martin Soler, from issue 171 of his newsletter, published in September 2026.
Start with context, not tools. Soler's core point: "AI needs context before it can meaningfully optimize anything." Before a hotel can use AI to improve operations, it needs a structured way to describe its own knowledge. That structure is an ontology, and it is less complex than the word suggests.
What an ontology means for a hotel. It is a description of what actually exists in the property, how those things relate to each other, and what can happen between them.
The Room 407 example. A single room carries more knowledge than any one system holds:
- A Deluxe King that connects to 408
- Takes 31 minutes to clean to standard
- Had its AC repaired two weeks ago
- Sits 75 metres from the main linen store
Layered on top is institutional knowledge. The executive housekeeper knows 407 often takes longer because of its layout, staff know which equipment performs poorly, and SOPs govern how the room is cleaned and with which products.
What AI can see once it is connected. Combine those static facts with live data (stock levels, arrivals, departures, rates) and patterns surface. Housekeepers might spend 18 minutes per shift walking back for supplies, or certain room sequences might run long because staff move between floors. Soler frames the payoff not as demanding faster work, but as making a physically demanding job a little easier.
Where to begin:
- List the core elements of the hotel: rooms, employees, tasks, equipment, and spaces
- Identify which systems already hold data about each
- Gather the SOPs
- Capture what is not documented and that "everybody knows"
- Use AI tools to structure it progressively, without collecting everything simply because you can
Why this matters for hotels. You do not need an "ontology platform" or a three-year data project. A modest, well-structured map of your property is the foundation that makes every later AI investment in operations more useful.
The same issue also covers Marriott supplying content directly to AI platforms, the history of the GDS as a toll booth that NDC added players to rather than removed, and why human travel advisors still matter in an agentic world.