Look at the open jobs in most maintenance systems and you’ll find an unhealthy distribution: a large pile marked high, a scattering of medium, and almost nothing low. This isn’t because everything is genuinely urgent. It’s because priority has been left as an opinion, and everyone’s opinion of their own job is that it matters.
Priority becomes useful the moment each level carries a defined response — an hour, eight hours, twenty-four hours, five days. Not because those particular numbers are sacred, but because a level with a number attached is a commitment, and a commitment is something that can be met or missed. “High” is a mood. “Respond within eight hours” is a fact you can measure against.
Two effects follow. The first is honesty at the point of raising a job: someone deciding whether a fault genuinely needs a one-hour response thinks harder than someone choosing between “high” and “medium”. The second is that performance becomes measurable — you can finally answer whether the estate is meeting its own standard, rather than whether people feel busy.
The most common mistake is grading by how annoying something is. A blocked sink in a public area is irritating and low-risk; a fire door that won’t latch is unremarkable to look at and a serious exposure. The scale should be anchored in what happens if the fault persists — to safety first, then to statutory position, then to operations, then to appearance.
It’s worth writing that hierarchy down, in one paragraph, and putting it where jobs are raised. Most priority inflation is not gaming; it’s absence of guidance.
The same applies to the lifecycle. A job list where everything is “in progress” tells you nothing. Useful status sets distinguish between not started, actively being worked, blocked or on hold, awaiting approval, deferred, and complete — because those states call for different responses. “On hold” needs a reason. “Awaiting approval” needs a name. “Deferred” needs a date, or it’s just a polite way of saying never.
One small mechanism does a disproportionate amount of good: generating a reference at the point the job is created. It gives the work a single identity that the finance system, the contractor and the compliance record can all refer to. When someone later asks what a particular invoice paid for, the answer is a job, which links to a finding, which links to a report. That chain is what turns “we spent it” into “here’s what it fixed.”
Once the scale means something, one figure becomes worth watching above the rest: the number of P1 jobs currently past their response time. Not the total open count, not the completion rate — just that. It’s small, it’s specific, and it’s very hard to argue with in a meeting.