The surveillance objection arrives before the pilot does — from housekeepers, from unions, from works councils. It deserves a straight answer rather than a reassuring one.
Every conversation about verified housekeeping reaches the same moment. Someone — a housekeeper, a shop steward, a works council member, sometimes the general manager on behalf of their team — asks the question plainly: so you want to watch us work?
It is the right question. A hotel that cannot answer it properly should not deploy anything, because the answer will be supplied by someone else, in a worse version, in the staff room.
Most of the fear in this conversation comes from imagining the wrong technology. So it helps to be concrete about what a motion-based system reads, and what it simply does not have access to.
Hotel IQ works from two signals. The first is motion — the accelerometer and gyroscope in a smartwatch or a phone, the same sensors that count your steps. An AI model recognises patterns in that movement and classifies them as cleaning activities: scrubbing, changing sheets, vacuuming, dusting, wiping, window cleaning. The second is Bluetooth proximity to a beacon in the room, which answers one question only: which room is being worked right now.
That is the entire input. There is no camera anywhere in the system. There is no microphone. There is no GPS trail of a person across the property or outside it.
"The subject of the record is the room, not the person. That is not a marketing distinction — it is the difference between proof of service and employee surveillance."
AiCarpus · Hotel IQThe honest concern behind the objection is rarely about data. It is about consequences: will this number be used against me?
A system built to rank people looks different from one built to verify rooms. Ranking needs comparable individual scores, a leaderboard, a target per person and a review cycle attached to it. Verification needs the opposite: a record attached to rooms, aggregated effort for the shift, and exception flags that point at a room where a phase is missing — not at a person who is slower than average.
This matters practically, not just ethically. A housekeeper who believes the watch is scoring them will optimise for the score — moving to look busy, skipping the pre-spray soak that makes a bathroom faster, rushing the rooms that get measured. You end up measuring the performance of measurement. A system that verifies rooms produces no incentive to game, because there is nothing to win.
In Europe, this conversation is usually formal, and it is better to arrive with answers than with a brochure. These are the questions that come up, in roughly the order they come up.
Only which room a cleaning session belongs to, and only while cleaning is detected. There is no location outside cleaning zones — not in the staff room, not on a break, not outside the building.
No. There is no camera and no microphone in the system, in any edition. The sensors involved are the motion sensors that count steps.
The product does not produce individual performance rankings. Effort metrics — steps, body load, work tempo, fatigue drift — exist to balance workload and flag overload, and are reported at shift level.
That is not a software question, it is a policy question — and it should be answered in writing before the pilot starts. Our recommendation to every property is the same: agree explicitly what the data may and may not be used for, and put it in the works agreement rather than leaving it to interpretation.
It belongs to the property, with a defined retention period and a GDPR mode that strips anything flagged as sensitive from every report. Purpose limitation is the point: data collected to prove a room was cleaned is used to prove a room was cleaned.
The mobile edition runs on the phone already in the pocket, so the wrist device is a choice about how much detail the property needs — not a condition of doing the job. Pilots run best on volunteers.
The reaction we see in real deployments is not enthusiasm on day one. It is scepticism for a few days, and then something closer to relief — for four reasons that have nothing to do with technology.
The people most likely to be harmed by an unverifiable housekeeping operation are the housekeepers. They are the ones who absorb the accusation, the re-clean, the unrealistic board, and the argument that ends with somebody must have skipped it.
"An objective record is not evidence against the person who did the work. It is usually the only evidence for them."
AiCarpus · Hotel IQThe failure mode here is procedural, not technical. A property announces a smartwatch rollout in a shift briefing, and by lunchtime the story on the floor is that management bought a tracking device. Four things prevent that:
Hotel IQ Mobile is the version to begin with in a privacy-sensitive environment: it runs on the phone the housekeeper already carries, needs no hardware, and produces the lightest record that is still useful — which rooms were serviced, how long each took, and whether the movement pattern matches cleaning.
Hotel IQ Watch adds phase-level verification per room when a property needs proof of service rather than proof of presence. Same engine, same privacy rules, one upgrade path — taken when the team already trusts the first step.
See the privacy model in the product: Explore Hotel IQ →
Related reading: When a guest accuses a housekeeper →
Hotel IQ verifies rooms, not people — no cameras, no microphones, no rankings, and a GDPR mode on every report. Book a demo and see exactly what gets recorded.