5 Maintenance KPIs Every Small Factory Should Track

Why five KPIs, not fifty

Maintenance management textbooks list dozens of metrics: OEE, MTTR, MTTF, wrench time, schedule compliance, planned maintenance percentage, mean time to acknowledge, cost per unit produced. For a plant with a dedicated reliability engineer, these are valuable. For a 30-person factory where the maintenance “team” is two people and the owner, most of these metrics require data you don't have and time you can't spare.

The five KPIs below give you a complete picture of your maintenance program's health. They're simple to calculate, easy to act on, and they answer the only questions that matter: Are we doing the PMs? Are things breaking less? Are we spending wisely? Is work piling up?

KPI #1: PM completion rate

What it measures: The percentage of scheduled preventive maintenance tasks completed on time.

How to calculate: (PMs completed on time ÷ PMs scheduled) × 100 = PM completion rate

What “good” looks like: 90% or higher. If you're below 80%, you effectively don't have a PM program — you have a list of things you meant to do.

This is the single most important maintenance metric for small teams. It answers the most basic question: are you actually doing preventive maintenance, or just planning to? A PM completion rate below 80% almost always correlates with high unplanned downtime, because the PMs that get skipped are the ones that prevent breakdowns.

Common reasons completion drops below target:

• Too many PMs scheduled — the program was built for an ideal world, not your actual staffing
• PMs take too long — checklists are bloated with unnecessary steps
• Reactive work crowds out PMs — every breakdown pushes a PM to “next week”
• Nobody is tracking it — if completion isn't visible, it's not prioritized

KPI #2: Mean time between failures (MTBF)

What it measures: The average time between unplanned breakdowns for a specific piece of equipment.

How to calculate: Total operating time ÷ Number of failures = MTBF

What “good” looks like: This varies wildly by equipment type. What matters is the trend. If your MTBF for the hydraulic press was 45 days three months ago and it's now 90 days, your PM program is working.

MTBF is your report card on reliability. A rising MTBF means equipment is failing less often — which is the entire point of preventive maintenance. A falling MTBF means something has changed: a PM is being missed, a part is wearing out faster than expected, or operating conditions have shifted.

For small shops, track MTBF on your five most critical machines. You don't need to track it on every piece of equipment. Focus on the machines where a failure stops production.

KPI #3: Downtime hours

What it measures: Total hours of unplanned equipment downtime per month.

How to calculate: Sum all hours where equipment was down due to unplanned failures. Don't include planned maintenance downtime.

What “good” looks like: Less than 2-3% of total available production hours. A single-shift shop running 20 days/month has 160 production hours — aim for under 5 hours of unplanned downtime.

Downtime hours convert directly to lost revenue. If your shop bills $150/hour and you lost 12 hours of unplanned downtime last month, that's $1,800 in lost production. Track this monthly and the pattern tells you whether things are improving or getting worse.

The key distinction is unplanned vs. planned. Shutting down a machine for a scheduled PM is planned downtime — it's an investment. A machine going down because a bearing seized is unplanned downtime — it's a failure. Only unplanned downtime goes into this KPI.

KPI #4: Parts spend

What it measures: Total monthly spending on maintenance parts and materials.

How to calculate: Sum all parts purchases tied to maintenance work orders — filters, belts, bearings, lubricants, replacement components.

What “good” looks like: Stable or declining over time. A sudden spike means something failed that shouldn't have. A gradual increase may signal aging equipment that needs capital replacement, not more maintenance.

Parts spend reveals two critical insights. First, it shows whether you're doing planned purchasing or emergency purchasing. Planned parts bought through your normal supplier cost X. Emergency parts bought from a distributor at 5 PM on a Friday cost 2-3X. If your parts spend is high, check how much of it is emergency purchases.

Second, parts spend by equipment tells you which machines are becoming money pits. If you spent $4,000 on parts for the 1998 lathe last year and the machine is worth $8,000, it's time for a capital replacement conversation — not another repair.

KPI #5: Backlog age

What it measures: How long open work orders have been sitting in the queue.

How to calculate: Average age (in days) of all open maintenance work orders.

What “good” looks like: Under 5 days average. If your average backlog age is over 14 days, work orders are being created and ignored — which means your team is either understaffed or the work order process is broken.

Backlog age is the canary in the coal mine. A growing backlog means maintenance demand is outpacing capacity. This happens gradually — one extra work order per week that doesn't get done, until suddenly you have 30 open items and no idea which ones matter.

The fix is usually prioritization, not headcount. Review the backlog weekly. Close anything that's been open for 30+ days and nobody has touched — if it hasn't been done in a month, it's either not important or the scope needs to be reconsidered. Focus on clearing the critical and high-priority items first.

How to actually track these

You can calculate all five KPIs from a spreadsheet if you log work orders consistently. The problem is that “consistently” almost never happens with spreadsheets. The data entry gets skipped when things are busy, which is exactly when the data matters most.

A CMMS calculates these metrics automatically from the work orders your team is already completing. Every time a tech marks a PM complete, the system updates your PM completion rate. Every time someone logs a breakdown, downtime hours and MTBF update. Parts used on work orders flow into parts spend. Open work orders age automatically.

The difference isn't the math — it's the data capture. If your system makes it easy to log work (phone, QR code, 2-minute workflow), you get accurate data. If logging work means walking to the office and opening a spreadsheet, you get gaps.

Start with one, then add

Don't try to track all five on day one. Start with PM completion rate — it's the most actionable and the easiest to measure. Once you're consistently above 85%, add downtime hours. Then MTBF. Then parts spend and backlog age.

The goal isn't a dashboard with perfect numbers. It's visibility into whether your maintenance program is actually preventing failures and controlling costs. Five numbers, reviewed monthly, give you that visibility without turning your maintenance team into data analysts.