What Metrics Actually Matter for a Maintenance Technician in Discrete Manufacturing?
Most performance conversations for maintenance technicians focus on the wrong level. You hear about plant OEE, departmental downtime totals, maintenance cost as a percentage of replacement asset value. None of those numbers are yours. None of them tell your Maintenance Manager specifically what you did or prevented.
There are three metrics that actually reflect individual technician performance in a discrete manufacturing environment. They are trackable without a dedicated analyst. They are meaningful to a Maintenance Manager in a performance review. And when you add one more number, the dollar value of failures you personally prevented, you have a contribution story that most technicians never build.
This guide covers those three metrics, what good looks like for each, how to track them simply, and exactly how to use them in a review conversation.
What Most Maintenance Technicians Get Wrong About KPIs
Waiting to be evaluated instead of tracking themselves. Most technicians let performance reviews happen to them. The Maintenance Manager looks at work order counts and asks how the month went. The technician who has been tracking their own metrics walks in with numbers, and the conversation is completely different.
Claiming credit for reactive work and nothing else. Fixing a stamping press motor after it fails is visible. Responding to an alert two weeks earlier, confirming a bearing fault, scheduling a planned repair, and preventing the failure entirely: that is invisible unless you document it. The technician who only gets credit for emergencies is also the technician who looks like the one always in the middle of emergencies.
Confusing busyness with impact. A high work order count does not tell a manager whether your work prevented anything. Three prevented failures with documented dollar values tells a clearer story than 40 completed work orders with no context.
Tracking the wrong unit. Technician performance is individual and asset-level, not departmental. Your MTTR on the stamping press motors you own is meaningful. The plant's average MTTR across all assets tells you nothing about your personal skill progression.
Metric 1: Response Time to Asset Health Alerts
When a condition monitoring system detects a developing fault on a stamping press motor or an assembly conveyor drive, it generates an alert. Response time is how quickly you move from that alert to a physical inspection and documented disposition.
What it measures: Your ability to act on information before a situation becomes an emergency. Fast response time is the difference between catching a bearing fault two weeks before failure and arriving at an unplanned breakdown with no parts staged and no context about what failed.
What good looks like:
- Severity-1 alert (imminent failure risk): inspection and disposition documented within 2 hours
- Severity-2 alert (developing fault, time-remaining measured in weeks): acknowledged and scheduled within 4 hours, inspection completed within 24 hours
- Severity-3 alert (early-stage trend, monitor): logged and added to next inspection round within the week
How to track it simply: For every alert you receive, note the alert timestamp and the time you completed inspection and documented your finding. You do not need a dedicated tool. A running log in a notebook or a recurring note in your CMMS work order is enough. At the end of each month, count how many alerts you responded to within target time versus outside it.
One sentence for your performance review: "This quarter I responded to [X] asset health alerts with an average response time of [Y] hours, documented findings on all of them, and escalated [Z] for planned repairs before they became emergency events."
The response time number matters. The documented finding is what turns a fast response into a prevented failure. You need both.
Metric 2: PM Completion Rate in Changeover Windows
Discrete manufacturing plants run on defined maintenance windows: model changeover shutdowns, weekend turns, holiday dark weeks. These are the planned opportunities to do the preventive work that keeps assets running between production runs.
PM completion rate is the percentage of your assigned scheduled tasks that you actually complete during those windows.
What it measures: Whether planned work is getting done or being deferred. Deferred PMs do not disappear. They accumulate in the backlog and show up later as unplanned failures on the exact assets that missed their service window.
What good looks like:
- 90% or higher completion rate in any given changeover window
- Any window below 80% should have a documented reason: emergency repairs displaced scheduled work, parts were not staged on time, scope was added without removing other scope
How to track it simply: After each changeover window, count your assigned PM tasks versus completed PM tasks. Note any that were deferred and why. If emergency repairs consumed your window, document that explicitly. It shows the constraint was upstream, not your execution.
Why the documentation matters: A Maintenance Manager seeing 70% completion with no explanation draws one conclusion. The same 70% accompanied by "three emergency repairs displaced four scheduled PMs; I completed the highest-criticality PMs and deferred the secondary conveyor lubrication tasks" tells a different story entirely.
One sentence for your performance review: "My PM completion rate in changeover windows this quarter was [X]%; in the two windows where completion fell below 90%, emergency repair displacement was documented in the work order history."
Metric 3: MTTR on the Assets You Own
Mean time between failure is a metric your Maintenance Manager watches at the asset level. Your version of that is MTTR, Mean Time To Repair, on the specific assets you are responsible for.
What it measures: Whether your repairs are getting faster and more effective over time on the asset classes you work on. A declining MTTR on stamping press motor bearings means you are getting better at diagnosing that failure mode, have the right tools ready, and are not wasting time troubleshooting from zero each time.
How to calculate it: Total repair hours divided by number of repair events on your specific assets over a rolling 90-day period. Track this at the asset class level, not across every work order you touch.
For example: if you completed 6 repairs on stamping press motor bearings in the last 90 days totaling 14 hours, your MTTR for that asset class is 14/6 = 2.3 hours. If it was 3.1 hours in the previous 90-day period, that improvement is a data point worth noting.
What good looks like:
- Consistent or declining MTTR on the asset classes you own
- No repeat failures on the same asset within 30 days of your repair (this is a quality indicator: the repair held)
One sentence for your performance review: "My MTTR on stamping press motor repairs dropped from [X] to [Y] hours over the past two quarters, with no repeat failures within 30 days of repair."
Benchmarks at a Glance
| Metric | Strong | Acceptable | Needs attention |
|---|---|---|---|
| Alert response time (Severity-1) | Under 2 hours | 2 to 4 hours | Over 4 hours |
| Alert response time (Severity-2) | Under 4 hours | 4 to 8 hours | Over 8 hours |
| PM completion rate per window | 90%+ | 75 to 89% | Below 75% |
| MTTR trend (rolling 90-day) | Declining | Stable | Increasing |
| Repeat failures within 30 days | 0% | Below 5% | Above 5% |
The One Number That Makes Your Contribution Undeniable
The three metrics above are operational. They show you are doing your job well. This number shows what your job prevented, and that is the number that moves a performance review conversation from evaluation to recognition.
The formula:
When you respond to an alert and confirm a developing fault before failure:
(Production value per hour on that line) x (Estimated hours to failure if undetected) + (Emergency repair premium vs. planned repair cost) = Your personal impact estimate
Here is a concrete example. You receive an alert on a stamping press motor bearing. You inspect it, confirm a developing fault, and schedule a planned repair during the next changeover window three weeks away.
If the bearing had run to failure:
- Line stop during production: 4 to 6 hours to diagnose, source parts, repair
- Production value lost: $10,000 per hour x 5 hours = $50,000
- Emergency repair premium: $6,000 vs. $2,000 planned repair cost = $4,000 premium
- Estimated consequence avoided: $54,000
That is one alert. One inspection. One documented finding. One scheduled repair.
How to document it simply:
- Asset name and ID
- Alert date and severity
- Fault confirmed (yes/no, describe)
- Repair scheduled (date, planned vs. emergency)
- Estimated consequence avoided (use the formula above with your plant's production value per hour; ask your Maintenance Manager for this number if you do not have it)
You do not need to be precise to the dollar. An estimate with visible reasoning is credible. A blank field is nothing.
One sentence for your performance review: "This quarter I responded to [X] alerts, confirmed [Y] developing faults, and documented approximately $[Z] in avoided production loss and emergency repair costs."
How to Use These in a Performance Review
You do not wait for the review to build this. You build it throughout the quarter.
After every alert response: log the time, the finding, the action, and the estimated consequence avoided. Takes two minutes.
After every changeover window: note your completion rate and any deferrals with reasons.
At the end of each month: calculate your running MTTR for the asset classes you own.
At the review: you have four numbers. Response time, PM completion rate, MTTR trend, and total estimated prevented cost for the quarter. You are not asking your Maintenance Manager to evaluate you based on impressions. You are presenting a record.
The technician who walks into a review with that record is not the same as the technician who walks in and waits to be assessed. And over time, that difference is what determines who gets the Reliability Technician conversation and who stays in the same role indefinitely.
How Tractian Gives You These Numbers
Tractian delivers the alert infrastructure that makes this possible. When a stamping press motor shows early vibration deviation or a conveyor drive bearing signature changes, the alert arrives at the technician level with the asset identified, the failure mode described, and the severity graded.
Your response time is logged. Your findings are documented against a work order. Your MTTR is calculable from your completed work orders. The estimated consequence avoided is a calculation you can make with the alert date, the repair date, and your plant's production value per hour.
The KPIs above are not aspirational. They are what condition monitoring infrastructure makes trackable at the individual technician level. The technician who uses this data to tell their own story is the one who advances.
See how Tractian supports maintenance technicians in manufacturing
Tractian continuously monitors equipment health in real time, detecting faults early and preventing unplanned downtime.
Explore the PlatformWhat metrics matter most for a maintenance technician in discrete manufacturing?
Three metrics define strong individual technician performance: alert response time (how quickly you act when the system flags a developing fault), PM completion rate in changeover windows (how much planned work you actually complete versus defer), and MTTR on the assets you own (whether your repairs are getting faster and more effective). The fourth number, the estimated dollar value of failures you prevented this quarter, is what makes your contribution visible to management.
What is a good response time to a condition monitoring alert?
Under four hours for a severity-2 alert on a production-critical asset. Under two hours for a severity-1 alert. Response time means the alert is acknowledged, the asset is physically inspected, and a disposition is documented, not just that the notification was read. Fast response paired with a documented finding is what creates a traceable prevented-failure record.
How do you calculate MTTR as a technician?
Total repair hours divided by number of repair events on your specific assets over a rolling 90-day period. Track this at the asset class level, not across all work orders. A declining MTTR on stamping press motor bearings means you are getting faster and more effective on that failure mode, exactly the skill progression that justifies advancement to Reliability Technician.
How do I put a dollar value on a failure I prevented?
When you respond to an alert and confirm a developing fault: (production value per hour) x (estimated hours to failure if undetected) plus the emergency repair premium versus a planned repair cost. A stamping press motor bearing caught two weeks early might represent $50,000 in avoided production loss plus $4,000 in avoided emergency repair premium. Document the alert date, confirmed fault, repair, and estimated consequence avoided.
What is PM completion rate and why does it matter?
The percentage of your assigned preventive maintenance tasks actually completed during available windows. Low completion means deferred work accumulates in the backlog and reappears as unplanned failures on the exact assets that missed their service window. Target 90% or higher. When emergency repairs displace planned PMs, document the reason. That record shows the constraint was upstream, not your execution.
How often should I update these metrics?
Alert response time: log every event when it happens. PM completion rate: calculate after each changeover window. MTTR: update monthly on a rolling 90-day basis. Prevented-failure dollar estimate: add each event as it occurs. Do not save this for review season. By the time the review arrives, you want three months of data, not a memory exercise.