If a maintenance organisation could track only one number to know whether it is healthy, the planned maintenance ratio (PMR) — also called planned-to-reactive ratio — is the one. It is the single best leading indicator of whether a maintenance team is in control of its plant or being controlled by it. And unlike most KPIs, it is dead simple to calculate and almost impossible to game without actually improving.
Here is what PMR measures, what a good number looks like, and why moving it is the whole game.
What the Planned Maintenance Ratio Is
PMR is the fraction of maintenance work that was planned in advance, versus the fraction that was reactive (unplanned, breakdown, emergency). It is usually expressed as a percentage:
PMR = Planned work hours divided by total work hours, times 100.
So a team that spends 800 hours on planned work and 200 hours on reactive work in a month has a PMR of 80%. The remaining 20% is reactive. World-class reliability programs run PMR in the 80 to 90% range; the average plant sits somewhere around 50%, and a plant in trouble runs under 30% — almost entirely firefighting.
A close cousin is schedule compliance — of the work that was planned and scheduled for the week, what fraction actually got done. Schedule compliance measures whether the plan was realistic and whether the team was allowed to execute it. PMR measures whether the work was planned in the first place. Track both.
Why PMR Is the Leading Indicator
Reactive work is expensive in every dimension. A breakdown repair costs three to ten times what the same repair costs when planned — emergency parts premiums, overtime labour, lost production, secondary damage from the failure, and the chaos tax of dropping everything to respond. Planned work is cheaper, safer, faster, and higher quality, because the parts are staged, the procedure is written, the isolation is arranged, and the right people are available.
A high PMR means the team is doing the cheap, controlled version of most of its work. A low PMR means the team is doing the expensive, chaotic version of most of its work. The ratio is a direct measure of how much money and disruption the plant is wasting on the difference.
More importantly, PMR is a leading indicator. Downtime, MTBF, and maintenance cost are lagging — they tell you what already happened. PMR tells you what is about to happen, because a rising reactive ratio today means rising downtime and cost next month. Move PMR up and the lagging metrics follow, with a delay. Watch PMR fall and the failures are coming.
What Drives PMR Up
PMR is raised by two things: doing more of the work in a planned way, and preventing the reactive work from happening at all.
Plan more of the work. The corrective work that is not an emergency — the "we know it needs doing but it is not urgent" jobs — should be planned, scheduled, and executed on the team's terms, not the machine's. This means a planning function that turns identified work into job plans (procedure, parts, labour estimate, tools, permits) ahead of time, and a scheduling function that slots those plans into the available labour windows. (This is core to lean maintenance — the waiting and chaos wastes are exactly what planning eliminates.)
Eliminate the reactive work at the source. The best reactive work order is the one that was never raised, because the failure never happened. Every preventive task, every condition-monitoring-driven intervention, every defect eliminated is a future reactive job converted into a planned one or removed entirely. Raising PMR is therefore inseparable from raising the reliability of the plant — the two are the same metric seen from different angles. (See defect elimination and preventive maintenance KPIs.)
The trap is the team that tries to raise PMR by planning the reactive work after it happens — writing a job plan for the breakdown repair so it technically counts as planned. That is gaming the metric and it improves nothing. Real PMR improvement comes from shifting work from the reactive bucket to the planned bucket before the failure forces it.
The 80% Target and the Last 20%
World-class is 80 to 90% planned. Getting from 50% to 70% is mostly planning discipline — a functioning planner-scheduler role, a job-plan library, weekly schedule meetings that are honoured. This is achievable for any plant within a year of focused effort.
Getting from 70% to 85% is harder, because the remaining reactive work is the genuinely unpredictable stuff — the rare failures, the odd failures, the things that are not yet caught by PM or condition monitoring. Pushing past 80% requires attacking those: root-causing the remaining failures (see defect elimination), deploying condition monitoring to catch failures before they become reactive (see oil analysis and vibration monitoring), and accepting that the last 10 to 20% will always include some genuine emergencies — the goal is to make them rare, not to pretend they can hit zero.
The plant that claims 95% or higher PMR is usually either redefining "planned" to include everything, or has so little preventive program that almost everything is emergency and the few planned jobs happen to dominate. Both are tells that the number is not real.
What Kills PMR Improvement
- No planning function. A team where every technician plans their own work on the fly has no PMR to speak of. Planning has to be a role, with time allocated to do it ahead of execution.
- Honouring the emergency over the plan. If every reactive job preempts the planned schedule, the schedule is fiction and PMR stays low. Leadership has to protect the planned work from interruption except for genuine emergencies.
- Backlog neglect. Identified-but-unplanned work piles up until it becomes emergency work. The backlog has to be actively planned and worked down, not ignored until it breaks. (See what is a maintenance backlog.)
- Measuring the wrong thing. Tracking "work orders completed" instead of "fraction planned" hides the ratio. The number has to be the ratio, reported monthly, trended over time.
How OpexMX Supports It
OpexMX computes PMR and schedule compliance live from work-order data — every closed work order is tagged planned or reactive at creation, so the ratio updates without anyone having to calculate it. The planner-scheduler view surfaces the identified-but-unplanned backlog so it gets converted to job plans before it ages into emergency work. And because condition-monitoring alerts (vibration, oil, thermography) open predictive work orders that count as planned, raising the monitoring coverage raises PMR automatically. The result is a PMR number the team can see, trust, and move — instead of a figure calculated once a year for a report nobody reads.