The Maintenance KPIs That Actually Matter for SMB Plants

Why Most SMB Plants Track Too Many Numbers (and Act on None of Them)

Picture this: the plant manager asks for a maintenance report at the Monday morning meeting. You open the spreadsheet — the one with eleven tabs, color-coded by someone who left two years ago — and spend twenty minutes assembling numbers that are already a week stale. By the time you read them aloud, two more PMs have gone overdue and the line has been down once.

This is the reporting loop that kills uptime. It is not a shortage of data. It is a shortage of the right data, surfaced fast enough to act on.

Most maintenance dashboards are built by adding metrics rather than subtracting them. Every corporate template or vendor demo throws in twelve to twenty KPIs, and the team dutifully populates them all — right up until the quarter gets busy and the spreadsheet stops being updated. A metric nobody acts on is noise. Noise costs time. And at an SMB plant with one or two maintenance staff, time is the scarcest input.

This guide cuts through the clutter. By the end, you will know exactly which five maintenance KPIs to track, what world-class benchmarks look like for each, how to calculate them from your own work-order history, and how to set up a single live view that drives actual decisions — without spending half your week in a spreadsheet.

---

The Five Maintenance KPIs That Drive Real Decisions

A useful maintenance KPI has three qualities: you can calculate it from data you already record, it tells you something actionable, and it moves when you change your maintenance behavior. The five below pass all three tests. They are not the only metrics in the field, but they are the right starting set for an SMB plant. Master these before adding anything else.

1. PM Compliance %

What it is. PM compliance % (preventive-maintenance compliance percentage) is the share of scheduled PMs that were completed on time in a given period.

Formula:

PM Compliance % = (PMs Completed On Time ÷ PMs Scheduled) × 100

What "on time" means. The practical standard — used by eMaint / Fluke Reliability — is the 10% rule: a PM is on time if it is completed within 10% of its interval. A monthly PM (roughly every 30 days) is on time if it is completed within about three days of its due date. A quarterly PM is on time within about nine days. The tolerance scales with the interval; the point is that "close enough" has a number, not just a feeling.

Benchmark. According to SMRP Best Practices (cited via eWorkOrders, 2026), world-class PM compliance is ≥90% overall, and ≥95% for critical A-class assets. Below 80% is the threshold at which a PM program is considered not functioning effectively — failures that look random often trace back to a compliance rate sitting in the 60s or 70s.

Why it matters first. Every other KPI in this list is downstream of compliance. MTBF cannot improve if PMs are not being done. Overdue aging is just a count of the compliance failures waiting to compound. Start here.

World-class target: ≥90% PM compliance overall; ≥95% for your critical assets. Below 80%, the PM program is not functioning effectively as a failure-prevention system. — SMRP Best Practices, cited via eWorkOrders, 2026.

For a deeper look at how to calculate and trend this metric, see the PM compliance % explained guide.

---

2. Planned-to-Unplanned Maintenance Ratio

What it is. The planned-to-unplanned ratio (sometimes called the planned maintenance percentage, or PMP) expresses what share of your total maintenance labor hours is spent on scheduled, proactive work versus reactive, break-fix work.

Formula:

Planned Ratio % = (Planned Maintenance Hours ÷ Total Maintenance Hours) × 100

Benchmark. Industry research aligned with SMRP guidance (Reliamag, 2026) puts the targets this way:

• Leaders: 90% planned / 10% unplanned
• World-class floor: 80% planned / 20% unplanned
• Reactive-heavy warning: below 70% planned

Most plants starting from a spreadsheet-based schedule sit well below 70%. That is not a character failing — it is a structural one. When there is no automated PM queue driving the day, reactive work fills every available hour because it announces itself loudly and the PM list does not.

Why it matters. This ratio is how you measure whether your PM program is real or aspirational. A compliance rate of 85% means little if only 30% of your labor is planned — it means you are completing a small, easy slice of PMs and spending the rest of the week reacting. The ratio exposes that.

What moves it. The surest levers are: reducing the backlog of overdue PMs (which become emergency reactive work), improving schedule adherence through earlier work-order generation, and recapturing wrench time lost to parts-hunting and administrative tasks. Industry research (Oxmaint, 2026) puts average wrench time at roughly 25–35% of a shift at most plants — meaning 65–75% of a technician's day is consumed by travel, waiting, documentation, and hunting. A planning-first approach — building the work package (parts, instructions, tools) before the PM is issued — is the structural fix.

---

3. MTBF — Mean Time Between Failures

What it is. MTBF (mean time between failures) is the average operating time between one failure and the next on a given asset. It is the core reliability signal for any piece of equipment.

Formula:

MTBF = Total Operating Time ÷ Number of Failures

Example: A compressor ran for 4,200 hours over the past year and failed three times. MTBF = 4,200 ÷ 3 = 1,400 hours per failure.

How to use it. Track MTBF per asset, not as a plant-wide average — a plant-wide MTBF number tells you almost nothing. The asset-level trend is what matters: is this pump's MTBF rising (the PM program is working), flat (holding steady), or falling (the asset is degrading faster than the PM program is catching)? A falling MTBF on a critical asset is an early warning that warrants a PM interval review, not a reactive repair on the next failure.

PM programs are documented to improve MTBF significantly — industry research (Re-Leased, 2025) summarizing reliability studies puts MTBF improvement at 50–75% with an effective PM program in place. That range is a general benchmark, not a guarantee; your results depend on your assets, duty cycle, and PM quality.

For the full calculation walkthrough, including how to handle partial operating periods and multiple failure modes, see the MTBF and MTTR calculation guide.

---

4. MTTR — Mean Time to Repair

What it is. MTTR (mean time to repair) is the average time it takes to restore an asset to service after a failure — from the moment the failure is detected to the moment the asset is back in production. It is your maintenance team's responsiveness and preparedness metric.

Formula:

MTTR = Total Repair Time ÷ Number of Repairs

Example: Three failures in a month required 2 hours, 5 hours, and 3 hours to restore service. MTTR = (2 + 5 + 3) ÷ 3 = 3.3 hours per failure.

What drives MTTR up. High MTTR is almost never a people problem — it is a process and parts problem. The three biggest drivers: (1) nobody documented the repair procedure, so the technician diagnoses from scratch; (2) the spare part is not on the shelf, adding hours or days of sourcing time; (3) the failure was not anticipated, so nobody is ready to respond. A good PM program attacks all three: documented procedures reduce diagnosis time, PM history surfaces which parts fail repeatedly so they can be stocked, and more frequent inspection catches degradation before full failure.

Benchmark. Industry research (Re-Leased, 2025) associates effective PM programs with MTTR reductions of 30–50%. Again, treat that as a directional reference tied to your own starting point, not a projection.

MTBF and MTTR are the two sides of the reliability coin — MTBF tells you how often you fail, MTTR tells you how long each failure hurts.

---

5. Overdue Task Aging

What it is. Overdue task aging is not a single number — it is a breakdown of every overdue PM or work order by how long it has been past due, typically bucketed into age ranges: 1–7 days, 8–30 days, 31–90 days, 90+ days.

Why it belongs on the dashboard. PM compliance % tells you your completion rate. Overdue aging tells you the severity of the gap. A plant with 85% compliance and no PM more than two weeks overdue is in a very different position than a plant with 85% compliance and six critical PMs sitting 60+ days past due. The aging distribution is where you find the next failure before it happens.

How to act on it. Sort overdue tasks by asset criticality first, then by age. The intersection of "A-class critical asset" and "90+ days overdue" is your fire. Every week, the question is not "how many tasks are overdue?" but "which overdue tasks on which assets pose the highest risk, and what is clearing them?"

A practical threshold. Any PM on a critical asset that has aged past one interval without completion warrants an immediate root-cause conversation — not a penalty, but a genuine "what broke in the schedule?" inquiry. Was the part not available? Was the technician pulled to a reactive emergency? Was the interval unrealistic for current staffing? The answer shapes the fix.

For a step-by-step breakdown of building and reading an overdue aging report, see the overdue task aging report guide.

---

One Bonus KPI for Plants Ready to Go Deeper: MC/RAV

Once the five core metrics are stable and tracked consistently, the next KPI worth adding is MC/RAV — maintenance cost as a percentage of replacement asset value.

Formula:

MC/RAV = (Annual Maintenance Cost ÷ Replacement Asset Value) × 100

Example: A plant spends $180,000 per year on maintenance (labor, parts, contractors) against a replacement asset value of $7,000,000. MC/RAV = ($180,000 ÷ $7,000,000) × 100 = 2.57%.

Benchmark. According to SMRP-aligned guidance (Factory AI, 2026), world-class plants run MC/RAV in the 2.0%–3.0% range annually. Below 2% can signal deferred maintenance — assets are not being maintained adequately, and failure debt is accumulating invisibly. Above 3% often signals a reactive-heavy culture spending more per repair than a planned program would cost.

MC/RAV is the only maintenance KPI that talks directly to finance in their language. It frames the maintenance budget not as overhead but as an investment rate against asset value — a framing that tends to unlock more productive conversations about PM staffing and parts inventory than any downtime horror story.

---

What World-Class Looks Like: A Benchmark Summary

KPI	Warning Zone	Functional	World-Class
PM Compliance %	< 80%	80–89%	≥ 90% (≥ 95% critical assets)
Planned-to-Unplanned Ratio	< 70% planned	70–79% planned	≥ 80% planned (leaders 90%+)
MTBF	Declining trend	Flat/stable	Rising trend per asset
MTTR	Rising trend	Flat/stable	Declining trend per asset
MC/RAV	> 3% or < 2%	~2–3%	2.0–3.0% of RAV

Sources: SMRP Best Practices via eWorkOrders (2026); Reliamag referencing SMRP (2026); Re-Leased industry research summary (2025); Factory AI SMRP-aligned guidance (2026).

These benchmarks are general industry targets — use them as orientation points, not pass/fail grades. Your specific assets, duty cycle, industry, and staffing ratio all affect where "good" sits for your plant.

---

How to Track These Maintenance KPIs Without Adding Administrative Work

The reason most SMB plants do not track maintenance KPIs consistently is not lack of intent — it is the administrative overhead of manual data collection. If calculating PM compliance % requires opening four spreadsheet tabs and cross-referencing two different lists, it will not get calculated every week. It will get calculated before the quarterly review, when it is too late to prevent anything.

The fix is structural, not motivational.

Option 1: A dedicated KPI spreadsheet. If you are not yet ready for a software platform, a well-designed Excel or Google Sheets dashboard can pull from your work-order log automatically — using lookups and pivot tables to produce the five core KPIs on a weekly refresh. The key is building the input structure correctly from day one: one row per work order, consistent date fields, a completion-status column, and an asset-criticality flag. Our Maintenance KPI Dashboard template is built on exactly this structure — pre-wired formulas, the benchmark thresholds already set, ready for your work-order history.

Option 2: A planning-first PM scheduling platform. A software system designed around the PM schedule — not the work-order queue — generates KPIs as a byproduct of normal operation. Every completed work order updates PM compliance % automatically. Every recorded failure updates MTBF and MTTR. Overdue aging is a live report, not a Friday afternoon calculation. The difference is not just convenience — it is that the data is current enough to act on before the failure, not after.

Maintenance Planning Manager is built on this planning-first model: define your assets and intervals, and the system generates the rolling PM work-order queue, calculates your compliance rate and overdue aging in real time, and surfaces your MTBF and MTTR from maintenance history — all on a flat-fee, unlimited-seat plan that does not add to the bill when you bring on a second technician. See the live KPI dashboard versus spreadsheet comparison if you want a side-by-side look at what that difference feels like in practice.

For full definitions and calculation notes on every metric mentioned here — and several more you will encounter as your program matures — the maintenance KPI glossary hub is the reference to bookmark.

---

The Metrics You Can Drop (For Now)

A brief note on what is deliberately left off this list, because the pressure to add metrics is relentless.

OEE (overall equipment effectiveness) — a powerful metric that combines availability, performance, and quality into a single production-efficiency score. World-class OEE is ≥85% (Oxmaint, 2026). It is also a measurement system that requires production data your maintenance team may not own or have clean access to. Start OEE tracking when your PM compliance and planned-ratio numbers are stable; layering OEE on top of an inconsistent PM foundation produces a number you cannot act on. When you are ready, the OEE guide walks through the full calculation.

Backlog hours — useful at scale, but at a two-person maintenance team, your overdue aging report covers the same territory with more actionable detail.

Cost per work order — meaningful for benchmarking and budgeting, but requires clean labor-hour tracking that most SMB teams do not have yet. Add it after the core five are running.

The discipline is not collecting fewer metrics because precision does not matter — it is collecting fewer metrics because the ones you actually review and act on are the ones that move the needle.

---

Start With One Number This Week

If you take nothing else from this guide: pull your PM compliance % for the last 30 days. Right now, from whatever system or spreadsheet you use. If you cannot pull it in under five minutes, that is itself the finding — the administrative overhead of your current tracking system is the first thing to fix.

The number you get will tell you exactly where to focus next. Above 90%? Move your attention to MTBF trends on your critical assets. Between 80–90%? Dig into which asset categories are dragging the rate down. Below 80%? The overdue aging report is your immediate priority.

The rest of the maintenance KPIs follow from there — each one sharpening the picture a little more, each one answerable from the same work-order data you are already (or should be) recording.

If you want a head start on the tracking infrastructure, the Maintenance KPI Dashboard gets the spreadsheet side ready in an afternoon. When you are ready to move beyond a spreadsheet entirely, Maintenance Planning Manager runs the calculations live, automatically, on a flat-fee plan built for the SMB team — one price, unlimited seats, no per-technician penalty for growing your crew.