CMMS reporting is where most products put their demo effort and where most sites get least value, because the reports that come as standard are mostly counts and the counts mostly cannot be acted on. Four reports do change decisions, and each one answers a question a maintenance manager is actually asked. This page is those four, with what each tells you and what to do when it moves, plus the most popular report in the category and why it is nearly useless.
Reactive share, quarterly, per site
The proportion of work nobody scheduled. On the worked example this site publishes it is 35%, worth 84 breakdowns and 193.2 avoidable hours a month, $9,273.60 at a $48 loaded rate. It is the honest health measure of a maintenance function and it moves slowly, which is what makes it trustworthy.
Backlog age, on the oldest open job
Not the count. Forty open jobs where the oldest is nine days is a working board; the same forty where the oldest is eleven months is a resourcing problem, and the count cannot distinguish them. Age is also legible to people outside maintenance, which matters when you need a decision from them.
Deferral, jobs rescheduled more than once
The earliest warning a maintenance function gets, and the one most products bury. Repeated deferral of the same jobs means either the interval is wrong for those assets or the schedule is over-committed against the technicians available, and both are fixable while they are still scheduling problems rather than breakdowns. Watch the repeat offenders rather than the total: five jobs deferred every month is a message, and forty deferred once between them is a bad month.
Repeat jobs per asset, rolling twelve months
Three of the same fault on the same machine is one unsolved problem, not three jobs. This report is what finds the two or three assets causing a disproportionate share of the work, and it is where the money usually is.
Questions people ask about cmms reporting
Why not PM completion percentage?
Because it is easy to hit and hard to learn from. A schedule at 95% with the same five jobs deferred every month is not at 95% in any sense that matters, and the metric hides exactly that.
How often should these be reviewed?
Backlog age and deferral monthly; reactive share and repeat jobs quarterly. More often than that and you are reading noise.
What should go to people outside maintenance?
Backlog age and reactive share, with the hours behind them. Both are legible and both lead to the resourcing conversation that maintenance usually needs to have.