Your PM Compliance Rate Is Lying to You (And What to Measure Instead)
A high PM compliance rate tells you that work orders got closed. It does not tell you whether the preventive maintenance was done on time, done thoroughly, or scheduled against the right interval — and those are the things that actually keep equipment running. A program can report 95% compliance and still be failing if those PMs were closed late, closed empty, or aimed at the wrong schedule. The numbers that reflect real protection are three: on-time completion within tolerance, PM backlog age, and the ratio of planned to reactive labor hours — the signals behind a fully optimized maintenance workflow.
What to measure instead of PM compliance rate:
1. On-time completion within tolerance — not just the percentage of PMs marked closed.
2. PM backlog age — how long overdue PMs have been sitting, not just how many.
3. Planned-to-reactive ratio — how much labor goes to planned work versus firefighting.
And read your close-outs for real data instead of only counting them.
6 Things a Healthy-Looking PM Compliance Rate Hides
1. The Work Is Being Logged From Memory, Not at the Asset
When closing a PM is slower than doing it — walk back to a shared desktop, log in, re-type what you already wrote on paper — technicians batch a week of work orders on Friday and fill in the rest from memory. Now your “complete” PMs are recollection, not record.
Fix: Let people close a PM where they’re standing, on the phone in their pocket, in the same two minutes they finished the task. eWorkOrders runs in a mobile browser with no app to download or update, so a technician can close out at the equipment instead of at a desk. Whatever system you’re on, if close-out isn’t happening at the point of work, your numbers are fiction.
2. “Complete” and “Done Well” Are Counted the Same
A PM marked done can mean a full inspection with readings and a noted follow-up, or a box ticked from the seat of the truck. Compliance rate treats them as identical. The tell is empty close-out: no meter reading, no parts used, no notes, no “found this, fixed that.”
Fix: Don’t count completed PMs — read them. Require one real data point on the PMs that matter, a reading, a condition, or a photo, and watch how fast the honest completion rate drops below the reported one.
3. A PM List Nobody Trusts Gets Rubber-Stamped
Most PM schedules are set once, off the OEM manual, and never touched again. Over time the intervals drift out of line with how the equipment actually behaves. We had a gearbox on a monthly lubrication PM that our own history showed was failing between weeks two and three — the schedule was “compliant” and the asset was dying anyway. People stop believing the list and start closing tasks just to clear the queue.
Fix: Pull your failure history, set intervals against real mean time between failures instead of the manual, and prune the PMs that never find anything. Our guide to building a preventive maintenance schedule that actually gets followed walks through the interval math.
4. Every PM Competes With a Breakdown for the Same Wrench
You have a fixed number of labor hours, and a breakdown always wins the argument. When reactive work spikes, PMs are the first thing quietly deferred — but if someone back-dates the close-out later, the compliance rate never flinches.
Fix: Track the planned-to-reactive ratio. If emergency work orders are climbing, your PMs are being sacrificed no matter what the percentage says. Measure how many labor hours actually went to planned work this period versus firefighting.
5. You’re Collecting Data Nobody Reads
Teams capture meter readings, failure codes, and close-out notes for years and never look at them. The compliance rate gets reviewed in a weekly meeting; the findings underneath it get reviewed by no one. So the same bearing fails four times and nobody connects the dots, because the connection is sitting inside closed work orders.
Fix: Pick a small set of numbers you’ll actually look at on a schedule — PM backlog age, repeat failures by asset, planned-versus-reactive hours — and put them in front of a human at a set interval. Data you never review is just storage.
6. A Compliance Number Is Only as Honest as the Data You Started With
A lot of lying compliance rates were born lying, at go-live. If assets got imported with duplicate names, missing serials, or PMs mapped to the wrong equipment, every rate you calculate afterward is built on sand.
Fix: eWorkOrders handles data migration during onboarding for exactly this reason — reviewing the existing records, standardizing naming and asset categories, and test-migrating one asset category before the full load. However you get there, audit your asset hierarchy and PM mappings before you trust a single percentage that comes off them. It’s one of the most common misses in a CMMS implementation.
Quick Reference: What the Compliance Rate Hides
| What looks fine | What it hides | Measure instead |
|---|---|---|
| PMs marked complete | Work logged from memory hours or days later | Close-out captured at the point of work |
| High % completed | Empty close-outs with no readings or notes | Real data point per critical PM |
| “Compliant” schedule | Intervals drifted from real failure behavior | Intervals set against measured MTBF |
| Steady compliance % | PMs deferred to chase breakdowns | Planned-to-reactive labor ratio |
| Data being collected | Nobody reviewing the findings | Backlog age + repeat failures, reviewed on a cadence |
| Any rate at all | Dirty data imported at go-live | Audited asset hierarchy and PM mappings |
Healthy Targets to Anchor Against
For context on where the honest numbers should land: PM compliance completed within tolerance at 90% or above, planned maintenance at roughly 85–90% of total labor hours, and emergency work orders under 10% of all work orders. If your compliance rate is high but your planned-maintenance share is low, the program is still reactive — the percentage is just masking it. A fully optimized maintenance workflow keeps all three moving in the right direction at once.
The Short Version
Compliance rate measures whether boxes got checked. It says nothing about whether the work was timely, thorough, correctly scheduled, or even real. Watch on-time completion within tolerance, PM backlog age, and your planned-to-reactive ratio instead — and read the close-outs, don’t just count them. Those numbers are harder to fake, and they’re the ones that actually track whether your equipment is protected.
Frequently Asked Questions
Is a high PM compliance rate always good?
No. A PM compliance rate above 90% can still hide late close-outs, empty work orders, PMs deferred to chase breakdowns, and schedules aimed at the wrong interval. Compliance rate only confirms that work orders were closed, not that the work was timely, thorough, or correctly scheduled. Track on-time completion within tolerance, PM backlog age, and your planned-to-reactive ratio alongside it.
What should I measure instead of PM compliance rate?
Measure three things the compliance rate cannot fake: on-time completion within the compliance window, PM backlog age, and your planned-to-reactive ratio. Then read your close-outs for real data, rather than only counting completed PMs.
What is PM backlog age?
PM backlog age measures how long overdue preventive maintenance tasks have been waiting, not just how many are open. A 95% compliance rate reached by closing PMs weeks late offers far less protection than 90% compliance completed within tolerance. Rising backlog age is an early signal that PMs are being deferred even when the compliance percentage still looks healthy.
Why do PM schedules become inaccurate over time?
Most PM schedules are set once from the OEM manual and never revisited, so intervals drift out of line with how equipment actually behaves. Technicians notice pointless or mistimed PMs and start closing them just to clear the queue. Setting intervals against measured mean time between failures (MTBF) and pruning PMs that never find anything keeps the schedule trustworthy.
How does data migration at go-live affect PM compliance data?
If assets are imported with duplicate names, missing serial numbers, or PMs mapped to the wrong equipment, every compliance rate calculated afterward is built on bad data. Cleaning and standardizing records during onboarding, and test-migrating one asset category before the full load, prevents a dishonest metric from day one.
Originally shared on our community at r/eWorkOrders.
