If you've ever asked that question — or had it asked of you in a leadership meeting — this article is for you. The PM compliance KPI everyone reports is usually measuring the wrong thing. Here's what's actually broken, and what to track instead.

The KPI everyone reports

Open any maintenance dashboard, anywhere in the world, and you'll find a number called PM Compliance — usually expressed as a percentage, usually somewhere between 70% and 99%, almost always reported in green or amber.

The textbook definition is simple: PM compliance is the percentage of scheduled preventive maintenance work orders that are completed by the scheduled date.

PM Compliance = (PM03 work orders completed on time) / (PM03 work orders scheduled) × 100%

Stage-3 plants on the NJN maturity scale typically hit 85–95%. Stage-4 plants hit above 95%. Stage-5 plants sit close to 98%.

So far, so familiar. Now let me tell you why the number you're looking at is almost certainly lying to you.

Three lies hidden in the standard PM compliance KPI

Lie #1: "Completed" doesn't mean "done"

In SAP PM, a work order is "completed" when it's TECO'd (technically completed). For PM compliance reporting purposes, it's "completed on time" if TECO happens on or before the scheduled date.

Here's the problem: TECO is an administrative status change. It does not require that the actual work was done.

In the field, here's what happens. The scheduled date approaches. The PM is at risk of going overdue. The KPI is tracked weekly. The maintenance manager doesn't want a red number on Monday morning. So the planner walks over to the system, marks the PM as TECO'd, and writes "completed" in the closeout notes — even though the artisan never actually went to the asset.

This isn't fraud. It's pressure on a number that has been disconnected from the work.

The fix: Audit a sample of "completed on time" PMs every month. Pull the work order, check the actuals (labour hours captured, parts consumed, condition data recorded). If a PM was "completed" in 5 minutes with no actuals captured, it almost certainly wasn't done.

Lie #2: "Scheduled" can be moved

PM frequencies in SAP are managed by the planner. If a PM is at risk of overrunning its due date, the easy fix is to move the next due date forward — push it from week 12 to week 14, and your "compliance" magically improves.

Many CMMS systems allow this without any audit trail. Even where there's an audit trail, nobody reviews it.

The fix: Track PM frequency drift as a separate KPI. For each major asset class, compare the current scheduled frequency against the FMECA-derived original frequency. Anything that has drifted by more than 25% needs an explanation.

Lie #3: "PM compliance" is a single number

The standard KPI rolls all PM types together. PM03 (preventive), PM04 (shutdown), and even sometimes PM01 work that has been miscoded — all aggregated into one percentage.

Here's a typical example from an East African gold mine. The plant-wide PM compliance was 91% — looking healthy. When we broke it down:

The plant was missing nearly a quarter of its highest-criticality PMs. That's the maintenance equivalent of skipping the safety inspection on the brake system to make sure you wash the windscreen on time.

"Real reliability is multi-dimensional. The KPI suite needs to reflect that."

What to track instead

If you want a leading indicator that actually predicts reliability, here's what to put on your dashboard:

1. PM Compliance — by criticality class

At minimum, three numbers: A-class compliance, B-class compliance, C-class compliance. Targets should be different — Class A should be above 95%, Class C can sit at 90% or even lower.

2. PM Schedule Adherence

The percentage of scheduled PMs executed on the scheduled date — not just before the due date. This catches the "TECO it before it goes red" behaviour that the standard compliance KPI misses.

3. PM Effectiveness Ratio (PMR)

The percentage of PMs that detect a defect or trigger condition-based follow-up. Healthy PMR sits around 15–25%. Below 5% means you're over-maintaining. Above 40% means your assets are degrading faster than your PM frequency is catching them.

4. Defect Density Trend

The rolling 12-month number of defects identified per work order completed. If your PMs are working, this number trends down.

5. PM-driven Corrective Ratio

The percentage of corrective work orders (PM01) that originated from a PM03 finding. High = your PMs are catching things before they fail.

Where this fits in the framework: This article draws from Book 6 — Asset Performance Management, particularly Chapter 10 (KPI catalog), Chapter 11 (Performance vs Theory — twin lenses), and Chapter 13 (KPIs by Role).