The hard question about any PM programme is whether it is doing anything, and it is hard because the successful outcome is an absence. Nothing broke. That is unprovable in the short term, which is why programmes get judged on activity instead and why sites end up proud of a schedule that is preventing nothing. Four readings, all taken from your own records rather than from any benchmark, get much closer to the truth, and none of them requires software to compute.
Completion, and then timing
How much of the due work was done is the first reading and the easiest to game, because a visit ticked at the end of the month is completion without maintenance. The second reading, how close to its due date each visit actually happened, is the one that shows whether the plan is being run or being caught up with. A programme with high completion and terrible timing is a monthly scramble.
Findings: is the work finding anything
Count the visits that produced a finding. Zero findings over a long period on an asset class either means the equipment is in excellent condition or the checks are not looking at anything, and the two are distinguishable only by reading what the checks actually are. This reading is the one that tells you a checklist has become a ritual.
Breakdowns between visits, on assets that are on the plan
The reading that matters most and takes longest to accumulate. A failure on an asset with a current PM schedule is direct evidence that the interval or the checklist missed something, and each one is worth a specific conversation rather than a statistic. Sites that treat these as individual evidence improve their intervals; sites that treat them as a rate do not.
Questions people ask about maintenance pm programs
What is a good PM completion figure?
Published benchmarks exist and are worth very little without knowing how a site defines due, completed and PM. Your own trend over time is far more informative than any external number.
Should we count time as well as visits?
Hours spent on planned versus unplanned work is the most revealing single ratio a maintenance department can keep, and it needs nothing more than time recorded against the right queue.
How long before the readings mean anything?
Completion and timing are readable within a month. Findings need a quarter. Breakdowns between visits need a year or more, which is the argument for starting the record before you think you need it.