Leading vs Lagging Maintenance KPIs: What to Track and Why

The Measurement That Arrives Too Late

The repair bill lands on Friday afternoon. The press brake went down Wednesday, ran an emergency parts order Thursday, and by the time the technician clocked out Friday the work order was closed — but the quarter is already $14,000 over budget. You open the maintenance log and find the last PM on that machine was eleven weeks ago. The interval was six weeks.

That is a lagging indicator in action. The number that told you something went wrong arrived after the damage was done. Somewhere in your operation there was a leading indicator — a scheduled task that slipped past due, a PM compliance rate that had been quietly declining — and it went unread.

The practical difference between a maintenance program that catches problems early and one that only records them comes down to which KPIs get dashboard space. By the end of this article you will know which metrics are leading (they predict failure before it happens), which are lagging (they confirm what already occurred), and how to track enough of both to run a maintenance program that budgets accurately and breaks down less often.

---

The Core Distinction: Predict vs. Report

A leading maintenance KPI measures an activity or condition that influences future equipment reliability. It is observable before a failure event and, if it moves in the wrong direction, gives you time to intervene.

A lagging maintenance KPI measures an outcome — downtime, repair cost, failure count — that is only knowable after the event has already affected the operation.

Neither type is optional. Leading indicators tell you where to aim your effort. Lagging indicators tell you whether the aim is working.

The failure mode most maintenance managers encounter is tracking almost exclusively lagging indicators — downtime hours, repair costs, failure counts — and treating leading indicators as nice-to-have. The result is a dashboard that is always perfectly accurate about the past and almost useless for preventing the next event.

A balanced maintenance metrics dashboard carries three to five of each type, updated on a cadence that matches the speed at which problems develop on your floor.

---

The Leading Maintenance KPIs That Matter Most

PM Compliance Rate

PM compliance is the ratio of preventive maintenance tasks completed on time to the total tasks scheduled in a given period.

PM Compliance Rate = (PM tasks completed on time ÷ PM tasks scheduled) × 100

If your equipment fleet has 40 PM tasks scheduled in a month and 32 are completed within their due window, PM compliance is 80%. The target threshold varies by operation, but the direction is unambiguous: a compliance rate that trends downward over consecutive periods is a leading signal that deferred work is accumulating — and accumulated deferred work is the precondition for the kind of unplanned failure that drives reactive repair costs.

Reactive maintenance is materially more expensive than planned PM work. When you defer a $300 lubrication task for six weeks, you may be setting up a $4,000 bearing replacement. PM compliance rate is the KPI that makes that dynamic visible before the bearing fails. See our PM compliance tracking guide for how to document and report this metric.

PM compliance is also the KPI most directly under your control. You can raise it today by adjusting your interval schedule, reassigning technician time, or simply surfacing which assets have tasks overdue. No specialized sensor or IoT integration is required.

Scheduled Maintenance Ratio (Planned vs. Reactive Work)

This metric tracks the proportion of all maintenance work that was scheduled in advance versus work that was reactive (triggered by a failure or breakdown). Operations without digital maintenance systems average roughly 40%–55% of their work in reactive mode; operations using maintenance software typically bring that figure to 15%–20% (MapTrack, 2026).

Scheduled Maintenance Ratio = (planned work orders ÷ total work orders) × 100

Illustrative example: A plant logs 60 work orders in a month. 38 are scheduled PM tasks; 22 are reactive repairs. The ratio is 63% planned / 37% reactive — above average, but with meaningful room to move toward the 80%–85% planned range that characterizes well-run operations.

A rising reactive share is a leading signal even before it shows up in downtime or cost figures, because reactive work tends to compound: a reactive repair on one machine often crowds out the scheduled PM on another, which then fails reactively in the next period.

Work-Order Backlog (Age and Volume)

Backlog is the total count and age of open maintenance work orders that have not yet been completed. A growing backlog — especially one where average task age is increasing — is a leading indicator of PM depletion. If tasks are being created faster than they are being closed, the maintenance program is running a deficit that will eventually surface as unplanned downtime.

Track backlog by asset class and by technician capacity. A backlog concentrated on one or two high-criticality machines is more urgent than the same volume spread across low-criticality support equipment.

---

The Lagging Maintenance KPIs That Anchor the Program

MTBF — Mean Time Between Failures

MTBF = Total operating time ÷ Number of failures

MTBF (mean time between failures) is a lagging indicator because it is calculated from failures that have already occurred. It tells you how long, on average, a specific asset or asset class ran between unplanned stops. A declining MTBF trend over rolling quarters tells you that an asset is becoming less reliable — that leading indicators like PM compliance or lubrication frequency may have drifted.

MTBF is also the input you need to set a defensible PM interval. If your press brake has historically failed every 900 operating hours on average, a PM interval of 600–700 hours gives you a reasonable buffer. Without MTBF data, interval-setting is guesswork. See MTBF, MTTR, and OEE explained for the full worked calculation, or use the MTBF/MTTR/OEE Calculator Workbook to run the numbers on your own fleet history.

MTTR — Mean Time to Repair

MTTR = Total repair time ÷ Number of repairs

MTTR (mean time to repair) measures how quickly your team restores an asset to service after a failure. It is fully lagging — there is nothing to measure until after the event — but it is a useful diagnostic for the cost side of downtime. A high MTTR on a specific asset class suggests either a technician-skill gap, a parts-availability problem (20%–30% of downtime duration is tied to parts availability, per Oxmaint, 2026), or a documentation gap that slows diagnosis.

When MTTR climbs on machines that were previously fast to restore, that trend is worth investigating before it compounds into extended downtime events.

Maintenance Cost as % of Asset Value (MC/RAV)

MC/RAV = (Annual maintenance cost ÷ Replacement asset value) × 100

MC/RAV — maintenance cost as a percentage of replacement asset value — is the standard fleet-level cost KPI endorsed by the Society for Maintenance and Reliability Professionals (SMRP, via Fiix, 2022). It converts absolute maintenance spending into a rate that is comparable across facilities, asset classes, and time periods.

Benchmarks: world-class operations typically achieve 2%–3% MC/RAV; 3%–4% is a reasonable target for most manufacturers; figures above 5% are a warning signal that the asset base is aging, maintenance practices are reactive, or both (Tractian, 2026). Some guidance sets the world-class threshold at approximately 2% (Ginder, "Maintenance as a Corporate Strategy," via ReliaMag, 2026).

MC/RAV is lagging because it reflects costs already incurred. But tracked quarter over quarter, it becomes a diagnostic: a rising MC/RAV is often the first budget-level signal that PM compliance has been slipping. The maintenance cost as percentage of asset value guide covers how to calculate and benchmark this metric in detail.

OEE — Overall Equipment Effectiveness

OEE = Availability × Performance × Quality

OEE (overall equipment effectiveness) is a composite lagging indicator that combines uptime, throughput rate, and quality yield into a single percentage. The industry average OEE across sectors is approximately 60% (InfluxData, corroborated by LeanProduction/Fabrico, 2024); world-class is defined as 85% or higher (Tractian citing Nakajima/TPM, 2026). Discrete manufacturers average around 66.8% OEE (Godlan, 2025).

OEE is primarily a production metric, but maintenance drives the Availability component directly. A facility actively managing OEE as a primary KPI has been associated with up to 25% lower maintenance cost and 10%–20% throughput improvement over an 18-month period (McKinsey, via Cryotos, 2026). Whether OEE belongs on your maintenance dashboard depends on whether your operation tracks it at the machine level — if it does, it is the richest single signal for justifying PM investment to plant leadership.

---

Building a Balanced Maintenance Metrics Dashboard

A workable maintenance dashboard for an SMB manufacturer does not need a dozen KPIs. It needs a small set with clear ownership and update cadence.

KPI	Type	Update cadence	Who owns it
PM compliance rate	Leading	Weekly	Maintenance manager
Scheduled maintenance ratio	Leading	Monthly	Maintenance manager
Work-order backlog (age + volume)	Leading	Weekly	Maintenance manager
MTBF by asset class	Lagging	Quarterly (rolling)	Maintenance manager
MTTR by asset class	Lagging	Quarterly (rolling)	Maintenance manager
MC/RAV	Lagging	Quarterly	Plant/operations manager
OEE (if tracked at machine level)	Lagging	Monthly	Operations manager

The PM compliance rate and the work-order backlog get the shortest update cycle because they move fastest and respond most immediately to technician behavior. MTBF and MTTR move slowly — calculating them monthly on a small asset fleet can produce statistically noisy results; quarterly rolling averages are more stable. MC/RAV is a budget-period metric; quarterly is usually sufficient for SMB operations.

For a walkthrough of how these metrics feed into a plant-level fleet review, see the plant manager fleet review guide.

---

From Metrics to Intervals and Costs

Tracking these KPIs in isolation is less useful than connecting them to the two decisions they are designed to support:

1. When should I next maintain each asset? — PM interval, set from MTBF history and OEM guidance, verified by PM compliance trending.
2. What will maintenance cost this year across the fleet? — Annual cost projection built from per-asset labor hours × rate + parts, rolled up to fleet level, benchmarked against MC/RAV.

Both questions require persistent, multi-asset calculation — not a one-time estimate in a spreadsheet. An Excel workbook breaks down at ten or more assets because interval logic, compliance tracking, and cost rollup interact in ways that static cells cannot maintain reliably. The preventive maintenance interval and cost planning guide shows how to structure that calculation for a full fleet.

The Maintenance Cost and Interval Planner handles both: it stores each asset's PM interval (set in days, hours, or cycles), calculates next-due dates continuously, and rolls per-asset labor-and-parts cost estimates into a fleet-level annual cost figure — with MC/RAV benchmarking on Professional plans and above. That is the difference between a persistent calculation engine and a one-time widget: the numbers stay current as your fleet changes, without rebuilding the spreadsheet.

---

Next Steps

If you want to start with the lagging side — calculating MTBF, MTTR, and OEE from your existing repair history — the MTBF/MTTR/OEE Calculator Workbook walks through each formula with worked examples you can apply directly to your own data.

When you are ready to connect those reliability numbers to a persistent interval and cost plan across your full equipment fleet, the Maintenance Cost and Interval Planner offers a 14-day free trial — no credit card required. Start with as few as five assets, set PM intervals, and see a fleet-level annual cost estimate in the same session.

The metrics are only useful if the intervals and budgets they inform are kept current. That is the problem a persistent calculation engine solves.