In February, ACI World published the 2025 Airport Service Quality (ASQ) results. Close to 707,000 passengers answered at the gate, on the day they flew. For an airport operator, one sentence in the release matters more than the list of winners: airport ambience and cleanliness were the strongest drivers of overall satisfaction.
If cleanliness carries that much weight, it is worth asking how an airport knows how clean the washroom in pier B was yesterday at four in the afternoon.
Three sources of information, and none belongs to the operator
Terminal cleaning is usually done by a contracted provider. Information about whether the work was done comes from three places.
The provider’s report. The provider writes about itself. This needs no assumption of bad faith: delivery is simply measured by the party that is paid on the basis of delivery. The report may well be accurate. There is just nothing to check it against.
Your own inspection. Inspectors walk a hand-picked fraction of the floor. What they see, the data is good for. What they don’t see, there is no data on. The route also tends to be denser where it is easy to walk and thinner where most passengers pass through.
The ASQ score. This measures what ultimately counts, but it arrives quarterly and does not say which washroom, which day or which hour. By the time it lands, the quarter is closed and the contract fee has been paid.
All three are useful. Of the two parties to the contract, though, the operator is the only one not working from its own data.
What your own data means in practice
The starting point is replacing the fixed cleaning rota. A fixed rotation sends staff to every washroom at the same interval, whether fifty passengers have passed it since or five hundred. Passenger flow follows flights, banks and delays, not the clock.
So measurement starts from passenger flow and sensors: door counts, consumable levels, air quality. From these, each area gets a risk score showing where conditions are likely to be getting worse right now. The score signals in two directions: to the provider, to clean, and to the operator’s own inspectors, to show where it is worth going.
Cleaning staff confirm they are in the room with a QR code or an RFID badge, and measurement starts from that moment. The result is one record, written by the sensors and the badge, that neither party can edit afterwards.
There are no cameras in a washroom. It is the first question a works council asks, and the answer has to follow from how the system is built, not from a policy document.
Fitted to ASQ, not to a score of our own
Introducing a new metric also introduces a new argument about whether the metric is right. That is worth avoiding. The operator is already measured on ASQ, so the measurement has to fit ASQ.
In practice this means washrooms and circulation areas are handled separately, the way ASQ asks about them separately. For circulation areas the method is before-and-after condition plus analysis of activity. Washrooms need a different method, because use and privacy are different there. The output is a weekly compliance view instead of a quarterly surprise.
Where this usually breaks
A sensor drops out. If a door counter stops, the area’s traffic looks like zero and its risk score looks low. Missing data therefore has to show up as missing, not as clean. It sounds like a detail, and it is exactly the kind of detail trust gets lost on.
Badging becomes mechanical. If staff are paid for the scan, they will scan whether or not they cleaned. So the badge only proves that someone was there. Whether the work was done is shown by how the score moves afterwards and by targeted inspection. Inspection becomes exception-based: the inspector goes where the numbers don’t add up.
Staff think they are being measured. The system measures the area, not the individual. Say that at the start and then stick to it. If someone starts using it for individual performance reviews, people will scan at convenient moments and the data loses its value. We wrote about this at more length in the post on scanner data.
The provider pushes back. Understandably, since it used to write the report. But a shared record protects the provider too. If an area was dirty because traffic ran well above what the contract was sized for, the shared data shows that, rather than an argument.
What we are not promising
The direction is for the ASQ cleanliness score to go up while spend on cleaning and inspection goes down, because the work goes where it is needed. How large that change is at a given airport can only be told from that airport’s own baseline, which is why there is no percentage here.
Four questions are enough for a first step, and none of them needs system access. How much do you spend on cleaning a year? How many inspection hours go into it? What is your ASQ cleanliness score today? And how do you check today that the work was done? If half of these are not measured, that is already the first finding.
Optimum Intelligence builds what cleaning verification needs, from field capture (Track · OptiTrack) through one operational data layer (Understand · OptiSight) to a record of execution (Manage · OptiCheck). It needs no optimizer. If you want to see what your own washrooms would show, let’s talk.