Maintenance KPIs: 15 Metrics That Actually Matter (EN 15341)

Most maintenance dashboards track 40 numbers and improve none of them. The EN 15341 standard and the SMRP body of knowledge converge on a far shorter list. Here are the 15 maintenance KPIs that actually drive reliability, cost and schedule, each with a formula and a benchmark.

Why most maintenance KPIs are noise

Walk into ten maintenance departments and you will see ten different scorecards. Most measure activity, not outcomes: work orders closed, hours booked, calls answered. Activity metrics feel productive, but they rarely tell you whether your assets are getting more reliable or your spend is under control.

The right maintenance KPIs do. Two reference frameworks cut through the noise:

• EN 15341 (Maintenance Key Performance Indicators) is a European standard that defines a structured catalogue of economic, technical and organisational indicators built from a small set of base quantities.
• SMRP (Society for Maintenance and Reliability Professionals) publishes a complementary set of metrics (its 5.x reliability and 3.x work-management series) with consensus definitions and harmonised formulas.

Where the two overlap, you have something close to an industry standard.

A metric earns a place on your scoreboard only if it passes three tests:

• Outcome-linked — it reflects reliability, cost, throughput or control.
• Actionable — a specific person can change it.
• Stably defined — calculated the same way every period, so trends are real.

The 15 metrics below pass all three. We group them the way practitioners use them:

• Reliability: MTBF, MTTR, availability, OEE, failure rate
• Productivity & schedule: PM compliance, schedule compliance, wrench time, % planned vs reactive, backlog weeks
• Cost: maintenance cost as % of RAV, cost per unit output, % of cost that is reactive
• Materials: MRO inventory turns, stockout rate

Reliability KPIs: MTBF, MTTR, availability and OEE

These five answer the core question: are the assets up, and when they go down, how fast do they come back? They are the most cited maintenance KPIs and the foundation everything else builds on.

1. Mean Time Between Failures (MTBF)

MTBF is the average operating time between failures for a repairable asset. It is the single best gauge of inherent reliability.

Formula: MTBF = Total operating time / Number of failures.

Benchmark: there is no universal target; MTBF is asset-specific. What matters is the trend and how you compare against your own fleet and the original-equipment design life. Rising MTBF means your reliability program is working. Compute it from run hours and failure counts with the MTBF/MTTR calculator.

2. Mean Time To Repair (MTTR)

MTTR is the average time to restore a failed asset to service, from the moment of failure to the moment it runs again. It is a maintainability metric and a direct measure of repair effectiveness, parts availability and access design.

Formula: MTTR = Total repair (downtime) time / Number of repairs.

Benchmark: shorter is better, and like MTBF it is asset-class specific. Watch the gap between mean and worst-case repair times. The tail is where the lost production hides.

3. Availability

Availability is the proportion of required time an asset can actually perform. It is the natural bridge between MTBF and MTTR.

Formula (inherent availability): A = MTBF / (MTBF + MTTR). Operational availability further nets out logistics and admin delays.

Benchmark: world-class continuous-process plants run inherent availability above roughly 95%, with highly critical or redundant systems pushed to 99%+. Context matters more than the absolute number.

4. Overall Equipment Effectiveness (OEE)

OEE multiplies three losses into one throughput-quality number, and it is the metric that connects maintenance to operations.

Formula: OEE = Availability x Performance x Quality.

Benchmark: the widely cited JIPM/Nakajima figure puts 85% as a world-class discrete-manufacturing target (roughly 90% availability x 95% performance x 99% quality); a typical unmanaged line sits near 60%. Calculate it with the OEE calculator, and convert lost availability into money with the downtime cost calculator.

5. Failure rate (and the bigger picture)

Failure rate (lambda) is simply 1 / MTBF, the number of failures per unit time. It feeds reliability modelling.

With enough failure data, fitting a Weibull distribution tells you which failure mode you face (infant mortality, random, or wear-out) and lets you set life-based replacement intervals. The Weibull calculator estimates the shape parameter and B10 life from your event history.

Productivity and schedule KPIs

Reliability metrics report the outcome. This next group tells you whether your work-management process — which produces that outcome — is healthy. These are the leading indicators a maintenance manager can move week to week.

6. PM compliance

The percentage of scheduled preventive maintenance completed on time within its window (often a 10% rule, e.g. a 30-day PM done within +/- 3 days).

Formula: PM compliance = PMs completed on time / PMs scheduled x 100.

Benchmark: SMRP and most practitioners target 90%+, with 95%+ considered best practice. Below roughly 80%, your preventive program is effectively optional and reactive work will dominate. Standardised, repeatable instructions raise compliance; the PM procedure generator produces consistent task lists.

7. Schedule compliance

The percentage of work planned into the weekly schedule that was actually completed as scheduled. It measures the integrity of your planning and scheduling loop, not just PMs.

Formula: Schedule compliance = Scheduled hours completed / Total scheduled hours x 100.

Benchmark: a mature planning function holds 85-90%+. Low compliance usually means too much break-in work, a symptom of poor reliability upstream.

8. Wrench time (tool time)

The fraction of a technician's paid time spent actively performing maintenance at the asset, rather than travelling, waiting for parts, searching for information, or getting permits.

Formula: Wrench time = Hands-on maintenance time / Total available labour time x 100.

Benchmark: unmanaged reactive shops measure 25-35%; well-planned organisations reach 50-55%+. Wrench time is the strongest argument for investing in planning and a CMMS; the CMMS ROI calculator translates wrench-time gains into payback.

9. Percent planned vs reactive

The share of maintenance labour (hours or work orders) that is planned and scheduled versus unplanned breakdown work. This is arguably the single most diagnostic metric of maintenance maturity.

Formula: % proactive = Planned work hours / Total maintenance hours x 100.

Benchmark: the long-standing reliability target is roughly 80% proactive / 20% reactive (the 80/20 rule). Plants stuck below 50% proactive are firefighting and almost always spend more in total.

10. Backlog (weeks of work)

The total identified, ready-to-execute work expressed as weeks of available craft capacity. Backlog is a health indicator, not a problem in itself; a small, healthy backlog keeps crews planned.

Formula: Backlog (weeks) = Total backlog labour hours / Weekly available craft hours.

Benchmark: the practitioner sweet spot is 4-6 weeks total (and about 2 weeks of "ready" backlog). Near zero means you are reactive and over-resourced; double digits means work is rotting and reliability is decaying. Size yours with the maintenance backlog calculator.

Cost KPIs: spending the right amount on the right work

Cost metrics keep the reliability program honest. The goal is never the lowest maintenance spend; it is the lowest total cost of ownership. That usually means spending more on proactive work to avoid far larger losses from downtime and emergency repairs.

11. Maintenance cost as a percentage of RAV

Annual maintenance cost divided by the Replacement Asset Value (the cost to rebuild the plant new). This is the headline benchmarking metric in both EN 15341 and SMRP because it normalises across plants of different sizes and ages.

Formula: % RAV = Total annual maintenance cost / Replacement asset value x 100.

Benchmark: SMRP cites a typical best-practice range of roughly 2-3% of RAV for many industries, though it varies widely by sector (refining, mining and pulp/paper run higher). Above about 5% generally signals a reactive, inefficient operation. Compute and benchmark yours with the maintenance cost % of RAV calculator.

12. Maintenance cost per unit of output

Total maintenance cost divided by units produced (tonnes, cases, MWh, vehicle-km). It links maintenance spend directly to what the business sells and is excellent for internal trending.

Formula: Cost per unit = Total maintenance cost / Units produced.

Benchmark: no cross-industry figure exists. Manage the trend and segment by line or asset class.

13. Percentage of maintenance cost that is reactive

The cost twin of metric #9. Emergency and breakdown work typically costs 3-5x the same job done planned, once you load in premium parts, overtime, expediting and collateral damage.

Formula: % reactive cost = Reactive maintenance cost / Total maintenance cost x 100.

Benchmark: driving this below roughly 20-30% of total cost is a primary objective of any reliability-centred program. For replace-vs-repair and intervention-timing decisions, model the whole picture with the lifecycle cost (LCC/TCO) calculator.

Materials KPIs: the MRO storeroom

Spare-parts inventory is where a surprising amount of working capital and downtime risk lives. Two metrics keep it balanced.

14. MRO inventory turns

How many times per year you consume and replace the value of your maintenance, repair and operations (MRO) stock. Higher turns mean less cash tied up, but pushed too far they raise stockout risk on critical spares.

Formula: Inventory turns = Annual MRO consumption value / Average MRO inventory value.

Benchmark: MRO turns are intrinsically low because of insurance spares; many plants land around 1-3 turns per year, with leaner operations pushing higher. Judge it alongside service level, not in isolation. Set order quantities rationally with the spare parts EOQ calculator.

15. Stockout rate / parts service level

The percentage of parts requests that cannot be filled immediately from stock. It is the counterweight that stops you from optimising inventory turns into a downtime disaster.

Formula: Stockout rate = Stockout occurrences / Total parts requests x 100 (service level = 100 - stockout rate).

Benchmark: target a high service level (often 95-97%+) on critical spares specifically, prioritised by asset criticality rather than blanket-stocking everything. Rank what to protect with the asset criticality calculator.

The 15 maintenance KPIs at a glance

Use this as a one-page scorecard. Treat benchmarks as directional; your industry, asset age and production model shift the right targets.

How to build a scorecard that actually moves

Fifteen metrics is a menu, not a mandate. Picking and sequencing them matters more than tracking all of them at once.

• Balance leading and lagging. Lagging outcomes (MTBF, availability, % RAV) tell you where you are; leading process metrics (PM compliance, schedule compliance, % planned) tell you whether tomorrow will be better. Pair them so a worsening leading metric warns you before the lagging one drops.
• Define once, in writing. The fastest way to ruin a KPI is to change its definition. Anchor each one to its EN 15341 or SMRP definition and document what counts as a failure, a planned job, or on-time. Most cross-plant benchmarking disputes are really definition disputes.
• Match the metric to the role. Planners own backlog and schedule compliance; supervisors own wrench time and PM compliance; the reliability engineer owns MTBF and failure modes; the manager owns % RAV and the proactive ratio. A KPI with no owner is decoration.
• Start with the maturity triad. If you measure only three things, measure % planned vs reactive, PM compliance, and maintenance cost as % of RAV. Together they reveal whether you are proactive, whether your preventive program is real, and whether you are spending the right amount. The other twelve refine the picture.
• Connect KPIs to money. A metric that cannot be traced to downtime cost, capital or throughput will lose budget battles. Route improvements through the downtime cost and CMMS ROI calculators so reliability gains land as dollars.

The broader management-system context is ISO 55001 (asset management), which expects organisations to set objectives and monitor performance, and reliability methodologies such as RCM (per SAE JA1011) that decide which failure modes are worth managing in the first place. The KPIs above are how you prove those systems are working. Explore the full toolkit on the tools page.

Related free calculators

Frequently asked questions