Is your maintenance plan working? 3 KPIs to prove it

On paper, every maintenance plan is good: preventive tasks well spread out, frequencies defined, no corrective work expected. In practice, the question is different: is the plan being followed, and is it preventing failures? Three KPIs answer that with numbers.

All three come from work orders: how much of the work is planned, how much of what's scheduled gets done on time, and how much has to be redone. Together, they show whether the problem is in the plan, in the execution or in the quality of the work.

1. Planned vs. corrective

This measures how much of the team's effort goes to planned work (preventive, predictive, inspection) and how much goes to firefighting (corrective).

% corrective = corrective labor hours ÷ total maintenance labor hours

% planned = preventive, predictive and inspection labor hours ÷ total maintenance labor hours

Note that the math uses labor hours, not the number of work orders. A 12-labor-hour corrective job on a gearbox weighs much more than three 20-minute inspections. And labor hours are not the same as clock hours: two technicians working for 4 hours make 8 labor hours.

Example from a fictional month:

Type	Labor hours	%
Preventive	480	48%
Predictive and inspection	140	14%
Corrective	380	38%
Total	1,000

Planned work: 62%. The most common reference is 70% or more planned work, and world-class operations usually sit near 80%, with no more than 20% corrective. Run the math with your own numbers in the planned vs. corrective calculator.

2. Schedule compliance

Having 70% preventive work in the plan doesn't help if half of it never leaves the paper. Schedule compliance measures how much of what was scheduled got done on time.

Schedule compliance = scheduled work orders completed on time ÷ work orders scheduled in the period

Example: in the month, 120 work orders were scheduled and 93 were done on time. Compliance = 93 ÷ 120 = 77.5%. A usual reference is to stay above 90%.

When compliance is low, the most common reasons are:

• Corrective work "stealing" the technician who was assigned to the preventive task.
• Parts or materials that weren't staged on the day.
• Equipment not released by operations.
• Scheduling above capacity: scheduling 150 labor hours in a week when the team delivers 120. The backlog shows whether this is happening.
• Unrealistic time estimates: the preventive task that "takes 1 hour" always takes 3.

Two rules keep this number honest. Record the reason for every work order not completed, from a fixed list, so you know which cause weighs the most. And don't quietly push the work order's date: rescheduling without a record makes compliance look great and hides the delay.

3. Rework

Doing the job on time isn't enough if it has to be redone. Rework measures the quality of execution.

% rework = labor hours spent on redone jobs ÷ total maintenance labor hours

A typical case: during preventive maintenance, the technician replaces a pump coupling but doesn't check the alignment. Two days later, the bearing overheats, the pump stops and the job has to be redone, now as corrective work. All the labor hours of that second visit are rework.

Example: in a month with 1,000 labor hours, 46 were spent redoing jobs. Rework = 4.6%. The ideal is zero; a common reference accepts up to 5%.

The most frequent causes of rework are:

• Shallow diagnosis: what was broken got replaced, not what caused the breakdown.
• Wrong parts, parts reconditioned without control or of doubtful quality.
• No procedure for assembly, torque, alignment and clearances.
• Testing cut short: the equipment was released without running for the time needed.
• Rushing because of a high backlog.

For rework to show up in the KPI, the return work order has to be linked to the original one. Without that link, the return looks like a new failure. See more in what to record on a work order. And every rework case is a good candidate for a 5 Whys analysis.

Reading all three together

Each KPI on its own tells part of the story. The combination points to where you should act:

What you see	What it usually is
High corrective and low compliance	Corrective work is eating the plan. Protect the team doing critical preventive tasks and tackle repeat failures.
High compliance and high corrective	The plan is followed, but it doesn't prevent failures. Review what gets done in each preventive task, not just whether it gets done.
Low corrective and high rework	The plan runs, but it's poorly executed. Work on procedures, training and testing before release.
Low compliance and low corrective	A plan larger than capacity, or unrealistic time estimates. Adjust the schedule.
Everything within the reference	Time to look at reliability and cost per asset.

How to track them

• Close all three for the same period, preferably monthly, on the same labor-hour base.
• Show the numbers to the whole team on a visual management board. The people who generate the data need to see the results.
• Follow the trend for three to six months before changing the whole plan because of one bad month.
• Split by trade and by area when the team is large.

In taskfield, the type of each work order, each technician's hours and the link to the previous work order are recorded through the app, and the three KPIs come out ready by period, with no side spreadsheet.