Decision-Grade Maintenance KPIs for Manufacturing

Key Points

• Maintenance KPIs become operational and add optimal value only when they're tied to specific goals, action thresholds, and decision owners. This will not happen if they're borrowed from generic industry checklists.
• KPIs like PMP, MTBF, wrench time, and maintenance cost per unit are interdependent. However, their deeper value emerges when they're read as a connected system rather than tracked in isolation.
• The infrastructure behind KPIs, including data capture, system integration, and automation, determines whether the numbers are decision-grade or simply reportable.
• Every maintenance KPI has cross-role impact (regardless of whether it’s explicitly recognized), affecting plant managers, reliability engineers, maintenance managers, and technicians differently.

The report looked fine, but the plant floor didn't

A reliability engineer at a mid-size food processing plant pulls up the monthly KPI report. MTBF is holding steady, PMP looks acceptable, and availability is within range. But three weeks later, two critical assets fail within the same production cycle. This causes a spike in reactive costs, and the maintenance team is back in firefighting mode. 

The KPIs didn't predict it because they weren't structured to do so. Each metric was being tracked in isolation, compiled manually from separate systems, and reviewed after the operational window to act had already closed.

This article addresses that issue. Not whether your facility is tracking maintenance KPIs, but whether the KPIs you track are structured to produce the kind of insight your operation actually needs. 

Maintenance KPIs impact manufacturing at a systems level. They shape how decisions get made, how resources get allocated, and how teams across the maintenance function coordinate around shared priorities. But their value depends on three things that most KPI programs don't fully account for. 

1. How KPIs get selected 
2. How the KPIs connect to one another 
3. Whether the infrastructure underneath them can deliver the decision-grade data needed to capitalize on their potential value 
4. Whether the infrastructure underneath them can deliver at the speed the decisions require

What Makes a Maintenance KPI Operational

A maintenance KPI becomes operational when it connects a measurement to a decision that someone in the organization is responsible for making.

Every manufacturing facility tracks numbers. Work orders closed, hours logged, parts consumed, failures recorded. But tracking a number and using a KPI to drive a decision are structurally different activities. 

A metric becomes a maintenance KPI only when three things are set:

• A defined goal that it measures against
• A threshold that triggers action when the number moves
• A person accountable for responding

Without all three, the facility is just collecting data. It’s not actually managing performance.

This may seem trivial or overly simplified, but it matters. KPI selection is a strategic decision. Or, it should be. Too often, it lands in the ‘task’ basket along with all the other ‘chores’ that need to get done. 

KPIs determine what you pay attention to, or ignore

The KPIs a facility chooses to track determine what the organization pays attention to and, just as importantly, what it ignores. When teams adopt KPIs from generic industry checklists without connecting them to their specific operational priorities, the numbers get reported on schedule but don't change how anyone works. For example, when a reactive maintenance ratio appears in a monthly summary, but nobody adjusts the PM schedule because of it. Or, an availability number gets presented to leadership, but no one investigates the three assets dragging it down.

The deeper problem is that poorly chosen KPIs create a false sense of visibility. The dashboard is full, and all the reports are generated. But the decisions those numbers are supposed to inform are still being made on instinct, experience, or co-opted by whatever problem is loudest that morning.

Asset performance metrics only become useful when they're matched to the operational questions the facility actually needs answered. And those questions vary by plant, depending on asset criticality, production demands, workforce capacity, and the facility's position on its maintenance maturity journey.

This is why the most impactful KPI programs aren't the ones with the most metrics. They're the ones where every KPI on the dashboard exists because it answers a specific question that someone acts on. And what makes those programs even more effective is understanding that KPIs don't operate alone. Their value multiplies or diminishes depending on how they interact with each other and with the systems feeding them data.

How KPIs Work as a Connected System

The KPIs that drive the most value in manufacturing aren't the ones that score well individually. They're the ones that reveal patterns when read together.

Many KPI programs treat each metric as a standalone indicator. Planned maintenance percentage gets its own widget. Mean time between failure gets its own trend line. And availability gets its own target. Each one is reviewed independently, as if improving one has no connection to improving the others. 

When approached this way, it overlooks the most important aspect of maintenance KPIs in a manufacturing environment. It’s that they're interdependent. One KPI's movement creates downstream effects on the others, and the cascade is where the deep operational insight lives.

Example One: How one metric's decline triggers the next

Planned maintenance percentage is a KPI that serves as a leading indicator for the entire system and will be our first example. PMP measures the proportion of maintenance work that's planned versus unplanned. When it drops, it doesn't just mean fewer scheduled tasks got completed. It means reactive work is consuming capacity. 

Technicians who were supposed to execute preventive inspections are instead responding to breakdowns. In turn, that compresses the time available for planned work the following week, which means fewer inspections, which means developing faults go undetected longer, which means MTBF starts to decline. The initial PMP drop didn't directly cause the MTBF decline. But it set the conditions for it.

To illustrate this, we’re going to show how it rolls across the major roles that touch the maintenance program. However, astute observers will notice an even deeper interdependence revealing itself through this brief illustration. Not only do KPIs function as a system for deeper operational insight and downstream impacts. KPIs also have implications for each role within a maintenance program.

Impacts across Maintenance

• Plant Manager: A declining PMP signals that maintenance costs are shifting toward emergency spending, complicating budget forecasts and making capital planning less predictable.
• Reliability Engineer: Lower PMP means fewer data points from planned inspections, reducing the reliability of failure trend analysis and weakening confidence in asset strategy decisions.
• Maintenance Manager: Falling PMP compresses the team's schedule, pushing technicians into reactive mode and increasing overtime without improving outcomes.
• Technician. When PMP drops, the daily work shifts from structured tasks with clear procedures to urgent calls with incomplete information, reducing the quality of both the repair and the data captured during it.

Example Two: What MTBF actually responds to

MTBF functions differently but is equally connected. Mean time between failure measures how long, on average, an asset operates before it fails. A rising MTBF confirms that the maintenance strategy is working. A declining one is an early signal that something upstream has shifted, whether that's PM frequency, inspection quality, parts availability, or operating conditions. 

The point is that MTBF doesn't improve in isolation. It responds to the quality of everything that feeds it. A facility can't directly "fix" MTBF. What it can fix is the PMP, inspection rigor, parts staging, and diagnostic clarity, which collectively determine whether MTBF moves up or down.

Impacts across Maintenance

• Plant Manager: Improving MTBF directly supports uptime targets and production commitments, making performance reporting more defensible to leadership.
• Reliability Engineer: MTBF trends validate or challenge the current asset strategy, providing evidence for adjusting PM intervals or recommending capital replacements.
• Maintenance Manager: A stable or rising MTBF means fewer emergency call-outs and more predictable workloads, which improves team morale and supports retention during a period of chronic skilled-labor shortages.
• Technician. A rising MTBF means fewer middle-of-shift interruptions and more time spent on planned work with proper parts and procedures in hand, which directly affects job satisfaction and the quality of each intervention.

Reading PMP and MTBF together tells a story that neither metric tells alone. A stable PMP alongside a declining MTBF might indicate that the PM tasks themselves need to be redesigned rather than merely completed. A rising PMP alongside a flat MTBF might mean the planned work isn't targeting the failure modes that actually cause breakdowns. 

These are the kinds of operational insights that surface only when KPIs are treated as a connected system rather than a collection of individual scorecards.

The Infrastructure Behind Useful KPIs

The quality of a maintenance KPI is only as good as the data, systems, and automation feeding it.

Even well-chosen, well-connected KPIs fail to produce operational intelligence when the underlying data is inconsistent, delayed, or manually compiled. In fact, so many KPI programs quietly break down over this very issue. The metrics look reasonable in a monthly report, but the data behind them was entered hours or days after the work happened, aggregated from disconnected sources, and averaged across time periods that obscure the operational moments that actually mattered.

Example Three: Wrench time and the granularity gap

Wrench time exposes this infrastructure gap more clearly than almost any other maintenance KPI. It measures the percentage of shift hours a technician spends performing actual maintenance work, as opposed to traveling between assets, waiting for parts, searching for procedures, or completing paperwork. 

Industry research has consistently placed the average at 25-35%, while best-in-class operations reach 55-65%. This gap represents one of the largest untapped productivity opportunities in manufacturing maintenance.

And herein lies an infrastructure challenge. Measuring wrench time accurately requires systems that capture task-level timestamps, not just work order open and close. Without that granularity, wrench time becomes an estimate, and estimates don't drive the kind of surgical process improvements that close a 30-point gap.

Impacts across Maintenance

• Plant Manager: Low wrench time means the facility is paying for labor hours that don't produce maintenance outcomes, directly affecting cost efficiency and headcount justification.
• Reliability Engineer: When technicians spend less time on tasks, inspection quality and data capture drop, weakening the inputs that condition monitoring analysis depends on.
• Maintenance Manager: Wrench time below 35% signals structural inefficiencies in planning, parts staging, or information access that no amount of overtime will fix.
• Technician. Low wrench time is felt before it's measured. It shows up as time spent walking back for tools, waiting on approvals, or searching for asset history that should have been accessible at the point of work.

Example Four: Connecting maintenance cost to production output

Maintenance cost per unit of production presents a similar infrastructure demand. This KPI ties maintenance spending to production output, unlike treating it as a standalone expense line. It's more operationally meaningful than total maintenance cost because it adjusts for volume. A plant that spends more on maintenance but produces significantly more units per dollar spent is outperforming one with a lower total budget but worse cost efficiency.

However, producing this KPI accurately requires integrated data from both the maintenance management system and the production system. Many facilities don't have that connection, which means the KPI either doesn't exist or gets built manually once a quarter from exported spreadsheets.

Impacts across Maintenance

• Plant Manager: This KPI translates maintenance spending into a language that production and finance leaders understand, making budget conversations more productive and resource requests more defensible.
• Reliability Engineer: Rising cost per unit despite stable failure rates may indicate over-maintenance or inefficient preventive maintenance compliance intervals that need strategic review.
• Maintenance Manager: Tracking cost per unit reveals whether maintenance improvements are actually translating into better plant economics or just shifting costs between budget categories.
• Technician. When the cost per unit is visible and trending in the right direction, it validates the daily work. Technicians can see that what they do on the floor connects to outcomes the organization values, not just tasks completed.

A McKinsey survey of 100 senior maintenance leaders found that preventive maintenance tasks accounted for only 51% of overall maintenance activity, compared with the 70-85% typically seen in organizations with mature reliability programs. This gap doesn't usually trace back to a lack of commitment. It traces back to systems that can't surface the right KPIs to the right people at the right time, making it harder to calculate and act on maintenance KPIs automatically, rather than manually compiling them after the operational moment has passed.

How Tractian Turns Maintenance KPIs into Operational Intelligence

We’ve described KPIs connected to decisions, systems-level interdependence between metrics, and infrastructure that produces decision-grade data at the speed operations require. Programs like this are exactly what Tractian was built to deliver.

Tractian's maintenance management platform provides real-time dashboards that track MTBF, MTTR, PMP, maintenance backlog, availability, and work order completion in real time, not after it's compiled. AI-powered analytics go beyond displaying numbers to identify patterns, surface anomalies, and generate prescriptive recommendations that connect each KPI to the action it's supposed to trigger. 

What makes this infrastructure particularly effective is the source of the data. Tractian's condition monitoring platform and sensors feed vibration, temperature, ultrasound, and runtime data directly into the platform, enriching KPIs with real-time asset health intelligence. This is how metrics like MTBF and availability become leading indicators rather than lagging reports. The platform's Auto Diagnosis capability detects all major failure modes and automatically triggers work orders generated with prescriptive guidance, closing the loop between detection and maintenance execution. 

This closed loop is what transforms a KPI from a number on a dashboard into a system that drives reliability decisions at the operational level.

At the strategic layer, Tractian's Asset Performance Management module adds failure libraries, root cause analysis, and benchmarking at the self, intra-company, and industry levels. This is the context that turns KPI trends into asset strategy decisions, giving reliability engineers the evidence they need to adjust PM intervals, justify capital replacements, or defend program changes to leadership. 

And with OEE integration, production monitoring, and energy reporting, maintenance KPIs can be linked to operational output, enabling metrics such as maintenance cost per unit of production to be calculated natively within the same platform.

Tractian's individual products also integrate with existing systems, including SAP, Oracle, and Power BI, and can be adopted as standalone tools alongside a facility's current infrastructure. Whether used as a complete platform or incrementally, Tractian enhances the production, capabilities, and impact of maintenance KPIs at every level.

Learn more about Tractian's real-time KPI dashboards and condition monitoring integration to see how high-quality, decision-grade IoT data transforms your program into AI-powered closed-loop maintenance execution workflows.

FAQs about Maintenance KPIs

1. How do I choose which maintenance KPIs to track first?
   Start with KPIs that connect to your most pressing operational priority. If unplanned downtime is the primary concern, PMP and MTBF provide the most immediate visibility into whether your maintenance strategy is improving reliability. Add KPIs incrementally as your data infrastructure matures rather than launching with a comprehensive dashboard that nobody acts on.
2. What's the difference between a maintenance metric and a KPI?
   A metric is a measurement. A KPI is a metric tied to a specific goal, a threshold that triggers action, and a person responsible for responding. Without those three elements, the number gets reported, but doesn't change how anyone works.
3. How often should maintenance KPIs be reviewed?
   Real-time dashboards should be available for daily operational decision-making. Weekly reviews at the team level and monthly reviews at the leadership level help identify trends and adjust strategies before small shifts become systemic problems.
4. Can maintenance KPIs work without a CMMS?
   They can be tracked manually, but spreadsheet-based KPI programs introduce delays and data entry errors that undermine accuracy. Automated data capture at the point of work is what makes KPIs trustworthy enough to act on, and that requires a system designed for it.
5. How do maintenance KPIs connect to production performance?
   KPIs like maintenance cost per unit and OEE directly tie maintenance outcomes to production output. When maintenance improves MTBF and reduces unplanned downtime, the facility gains production capacity without additional capital investment.
6. What role does condition monitoring play in maintenance KPIs?
   Condition monitoring feeds real-time asset health data into KPIs such as MTBF and availability, turning lagging reports into leading indicators. Without condition data, KPIs can only describe what has already happened. With it, they can inform what to do next.