Predictive Management

“It should take about four hours”

Ask most cleaning contractors how long a building takes to service and you’ll get an estimate based on square footage and experience. It’s usually roughly right, and roughly right is exactly the problem: it hides everything interesting.

Which part takes the longest? Where does the time actually go on a bad night? Has this building been taking longer since the third floor was reconfigured, or does it just feel that way? None of those are answerable from a total.

And because they’re unanswerable, the conversation about scope becomes a negotiation between two impressions.

Phases, each with a target

The shift is broken into the phases the service actually has — the specific ones for that account, not a generic template. Each phase carries a time target set for that building.

While the shift runs, status is visible: on target, drifting, over. Green, amber, red. The cleaner sees it, and so does the supervisor covering the account.

That visibility is deliberate and it is what separates this from timing people. A timer the worker cannot see is measurement performed on them. A timer they can see is information they can act on — including saying, during the shift rather than after it, that something in this building takes longer than whoever wrote the target believed.

Targets that can go down

Most systems of this kind only ever ratchet targets tighter. Ours adjusts in both directions, and the downward adjustments are the ones that make it credible.

A building with a difficult layout, older fixtures or restricted access hours genuinely takes longer, and measuring it against a target derived from a modern open-plan floor produces one of two outcomes: people rush and quality drops, or people learn the number is theatre and ignore it. Both are worse than having no target.

So when the history shows a phase consistently over target, the first question is whether the target is wrong. Usually it is.

What the data is for

The shift summary shows target, actual, and where the time went. Across weeks that becomes the most useful input to an account review we have: it says, in numbers, where the written scope and the physical building disagree.

That conversation almost always ends in a change to the scope, the frequency or the staffing — not in a conversation about an individual. A contractor using phase timing to pressure people is using it against the thing it measures, because a rushed shift scores badly on inspections and shows up in client feedback two weeks later.

Phases are driven from CleanPilot, so starting and finishing one is spoken rather than typed — which is why the timestamps are real.