CMMS and Spare Parts KPIs: Benchmark Ranges and How to Read Them
Maintenance and spare parts KPIs from wrench time to asset data quality: how to measure each one, the lever that moves it, and which figures have a cited public source.
Maintenance KPIs are most useful as a set: wrench time, planned maintenance compliance, backlog, schedule compliance, spares coverage, storeroom health and asset data quality. Use the ranges below as reference points for spotting a gap worth investigating, not as targets, and hold each KPI against its own trend first.
Every figure here shows its basis. A superscript number links to a public source in the Sources list at the end, and the sentence states that source's scope, such as its year and the kind of organization it covers. A figure marked no cited source is kept exactly as originally published because no public source stating it was found, so treat it as illustrative.
At a glance
CMMS, EAM and spare parts KPI benchmarks at a glance
Ranges are as originally published unless a source is cited. Where the original used a word other than typical, such as world class or healthy, that word is kept in the cell. The DOE FEMP benchmarks come from facilities operations and maintenance guidance published in 2010, which attributes them to NASA in 20001. Not stated means the original gave no figure for that column.
Method: How the figures in this guide are sourced
A figure with a superscript number is stated by the cited source, listed under Sources with its publisher, date and access date. Several cited sources are government facilities guidance or surveys from earlier years, and the sentence says so. A figure marked no cited source is kept exactly as originally published because no public source stating it was found. It is illustrative, not measured industry data. Worked examples are plain arithmetic on the numbers shown.
Wrench time
Wrench time is the flagship maintenance productivity KPI, and the gap between typical and world class is stark. Most reactive plants run 25 to 35 percent no cited source, which means roughly two-thirds or more of paid craft time is lost to waiting and walking. World-class planned organizations reach 55 to 65 percent no cited source.
Reactive work is still common. In a NIST survey of U.S. discrete manufacturers using 2016 data, maintenance averaged 45.7 percent reactive, 31.8 percent preventive and 17.3 percent predictive5.
The lever is planning and kitting: staging parts and permits before the job starts recovers 10 to 15 points on its own no cited source. Treat any month under 40 percent as a signal that your planner-to-technician ratio, ideally 1 to 15, is too thin no cited source.
How to measure
Measure hands-on tool time as a share of paid craft time.
Track it monthly, on the same sampling method each time.
Record where the rest of the time goes, such as waiting for parts, travel or permits.
What moves it
Job plans written before work is released to the crew.
Kitting that stages parts and permits before the job starts.
Enough planners for the size of the crew.
Measure how much paid craft time reaches the tools. Wrench Time
Planned maintenance compliance
Planned maintenance (PM) compliance separates proactive plants from firefighters. Typical sits at 70 to 85 percent no cited source; world class holds 90 percent or better1, counted with no force-closed padding.
That world-class figure matches the U.S. Department of Energy Federal Energy Management Program (DOE FEMP), whose 2010 operations and maintenance guide sets the benchmark for PM actions completed over PM actions scheduled at above 90 percent1. The same table puts emergency maintenance at under 10 percent of total hours worked2. Both are facilities benchmarks the guide attributes to NASA in 20001.
The same guide describes continually top-performing facilities as under 10 percent reactive, 25 to 35 percent preventive and 45 to 55 percent predictive6.
The improvement lever is schedule discipline: protect PM hours from break-in work by reserving 20 to 30 percent of capacity for the unplanned no cited source. Watch compliance and wrench time together, because plants that hit 90 percent compliance almost always cross 50 percent wrench time within two quarters as reactive work drains out of the schedule no cited source.
How to measure
Count only PMs closed inside their grace window.
Exclude force-closed work orders from the completed count.
Track emergency hours as a share of all maintenance hours worked.
What moves it
A capacity reserve for break-in work, so PM hours stay protected.
Schedule discipline that treats a missed PM as an exception to explain.
A steady shift of work from reactive to preventive and predictive.
Backlog is a health KPI with a real target band, not a number to minimize to zero. Healthy total backlog sits at 4 to 6 crew-weeks; ready backlog should stay at 2 to 4 weeks no cited source.
Below 2 weeks you are overstaffed or under-inspecting; above 8 weeks the queue is decaying and jobs age past relevance no cited source.
How to measure
Express backlog in crew-weeks: estimated backlog hours divided by weekly crew hours.
Report total backlog and ready backlog separately.
Track the trend, not the snapshot, and flag any month where ready backlog grows faster than completions.
Schedule compliance and labor load tell you whether the week is realistic. Schedule compliance is work completed as scheduled, and typical is 50 to 70 percent no cited source.
For the strong end, the DOE FEMP benchmark for schedule compliance, hours worked on scheduled jobs divided by hours scheduled, is above 90 percent, from the 2000 NASA industry benchmarks in its 2010 guide3.
Labor load should sit at 80 to 100 percent of net capacity, never chronically above 100, which guarantees slipped PMs and overtime no cited source. If load runs above 110 percent for a month, you are borrowing from next week's PMs no cited source. The same DOE FEMP table sets maintenance overtime at under 5 percent of regular maintenance hours4.
How to measure
Divide hours worked on scheduled jobs by hours scheduled.
Compute labor load against net capacity, after vacation, training and a break-in reserve.
Track overtime as a share of regular maintenance hours.
What moves it
Honest capacity planning that nets out vacation, training and a break-in reserve.
Schedules built on real available hours rather than headcount.
Critical spares coverage is the availability KPI that protects uptime. World-class plants hold 95 to 99 percent coverage on critical SKUs; typical drifts to 80 to 90 percent, which is where surprise stockouts live no cited source.
Pair coverage with stockout downtime exposure so you spend on the right parts. That lets you push coverage up without simply inflating inventory across the board.
How to measure
Define which SKUs are critical and keep that list under change control.
Measure coverage on the critical SKUs only.
Estimate the downtime a stockout of each critical part would cause.
What moves it
Disciplined reorder points tied to lead time and usage.
Reorder point reviews whenever supplier lead times shift.
Spending weighted toward the parts with the highest downtime exposure.
Inventory turns and inactive stock reveal whether the storeroom is lean or a money pit. MRO turns are naturally low; 1.0 to 2.0 is typical and 3.0 is strong no cited source.
The sharper KPI is inactive stock, items unused in 24 months no cited source. World class keeps that under 15 percent of line items; typical storerooms carry 25 to 35 percent dead no cited source.
Watch the carrying rate too, holding it near 18 to 22 percent rather than the 30 percent that unmanaged storerooms drift toward no cited source.
How to measure
Divide annual issues by average inventory value to get turns.
Count line items with no issue in 24 months as inactive.
State the carrying rate as an annual share of average inventory value.
Data quality is the KPI beneath every other KPI, because bad hierarchy corrupts all of them. Asset hierarchy completeness should reach 95 percent or higher for parent, criticality and bill of materials (BOM) fields no cited source.
Anything under 80 percent means your coverage and backlog numbers are guesses no cited source. Treat a completeness gain as a prerequisite, not a nice-to-have, before you set aggressive targets on the metrics above.
How to measure
Check parent, criticality and BOM fields on every asset record.
Re-run the check quarterly.
What moves it
A one-time cleanup campaign.
A governance rule that no asset goes live without full attribution.
Tie the KPIs to a financial scorecard so improvement survives budget season. Productivity gains read clearly once they are converted into output per technician.
For example, a plant moving wrench time from 32 to 48 percent gets 48 / 32 = 1.5 times the hands-on time from each technician, which effectively adds half a technician's output per head.
Combine that with PM compliance climbing past 90 percent and downtime typically falls 20 to 35 percent no cited source.
Published studies bracket that range, but they measure different interventions. In an A.T. Kearney and IndustryWeek survey of 558 companies using a CMMS, cited by DOE FEMP from 1994 data, users reported on average a 20.1 percent reduction in equipment downtime and a 17.8 percent reduction in MRO inventory7.
DOE FEMP also cites independent surveys that put the average downtime reduction from a functional predictive maintenance program at 35 to 45 percent8. In NIST's 2016 data, U.S. discrete manufacturers that invested more heavily in preventive or predictive maintenance had on average 44 percent less downtime, a cross-sectional comparison rather than a before and after measurement9.
Review the scorecard monthly, hold each KPI against its target band, and pick the single worst-performing lever to attack next quarter rather than chasing all metrics at once.
Quantify recovered uptime and inventory reduction so reliability gains read as dollars. CMMS ROI
How to build your own baseline
Write down the definition of each KPI, including grace windows and what counts as scheduled work.
Fix the asset hierarchy first, so the other KPIs rest on sound data.
Collect several months of data before you set a target.
Compare against your own trend first and published ranges second.
Record what changed in the maintenance program whenever a KPI moves.