Blog
The guest checks in at three, the room is ready at half four, and nobody is at fault
Hotel operations are rarely slow because people work slowly. They are slow because departments cannot see each other's queues.
Published
Three in the afternoon, a city hotel lobby. Four groups are waiting at the desk. The receptionist knows the rooms are not ready; she does not know when they will be. She calls the housekeeping supervisor, who does not answer because she is on the twelfth floor. So the receptionist says "just a few more minutes", which both she and the guests know is not an answer.
Meanwhile, on the seventh floor, two rooms have been finished since half past two. Nobody has reported it, because the housekeeper who finished them moved on to the next room and will report the whole round at the end.
The problem is not speed. It is visibility
Nobody in that story is working slowly. The housekeepers are on schedule. The receptionist is doing her job. The supervisor is solving something else.
What is missing is a shared status in real time, and it is missing because the system records room status as something a person enters afterwards rather than something that happens as the work happens.
What the right system changes
Status updated where the work occurs. One tap on a phone when a room is finished, and the lobby sees it immediately. This is not advanced technology, and a great many hotels still do not have it.
Sequencing rooms against the arrivals queue. If the system knows two superior rooms are due at two o'clock, it should put those rooms at the front of the round rather than working through by room number.
Forecasting demand. How many late checkouts, how many early check-in requests, how much laundry on a given day — all predictable from bookings plus historical patterns, which turns staffing from guesswork into planning.
Where AI genuinely fits
In hotels the word covers two very different things.
The first is a guest-facing chatbot, which handles repeated questions well — breakfast times, directions, extra towels — and handles an unhappy guest badly. The good design hands over to a person the moment the tone shifts, rather than trying to resolve it.
The second is forecasting and pricing, which models really do well, given enough history and someone watching. Events with no precedent — an airport closure, a new festival in town — send a model wrong in directions that cost money.
The common mistake
Hotels that buy one large system to replace everything at once tend to find staff back on paper within three months. A system that covers everything is rarely good at anything, and changing it all at once leaves nobody standing on familiar ground.
Starting at the single sorest point and making it work well enough that staff ask for the next piece is slower on paper and faster in practice.
What GIPSIC does here
We build booking and space management systems, and systems that have to work on the phone in a front-line employee's hand rather than on a screen in an office.
We tend to ask something in the first meeting that rarely comes up: what phones do your housekeepers carry, and what is the signal like in the basement? Those two answers determine whether a system gets used more than any feature does.
If you would like to talk about your hotel's operations, get in touch.
About the author
Film — Wisit. A businessman who still does his own BA work more often than he probably should, and writes a fair bit of code, front and back. Runs two or three small businesses. Follows technology and business obsessively, in Thailand and everywhere else. Off the clock: physics, astronomy, DIY, and anything to do with networks. Music always on, though he cannot sing. Plays instruments anyway, badly. Plays a lot of sport, racket sports above all. Not much of a traveller by himself, but happy to take Mint anywhere in the world. A man who fears — sorry, loves — his wife. One flaw: he barely touches video games.
Written with Claude Opus 5