Maintenance and Reliability KPIs: A Practical Guide

Introduction

Maintenance and reliability KPIs turn day-to-day work orders, failure records, labour hours and production interruptions into evidence for better decisions. For maintenance managers and plant leaders, the objective is not to create a crowded scorecard. It is to understand whether assets are being maintained in a controlled way, whether reliability risks are reducing, and where limited people, time and budget should be directed.

A useful KPI system connects maintenance activity to operational outcomes. It helps teams distinguish a temporary surge in repairs from a persistent failure pattern, test whether planning discipline is improving, and discuss priorities with operations using shared facts. This is particularly important in oil and gas, utilities and process industries, where a single critical asset can affect safety, environmental exposure, product quality, throughput and energy use.

The most effective maintenance and reliability KPIs are defined consistently, calculated from trusted data, reviewed at an appropriate frequency and followed by specific improvement actions. Measures alone do not improve performance; the decisions made from them do.

Why maintenance and reliability KPIs matter

Maintenance performance cannot be judged solely by whether technicians appear busy or whether a plant has avoided a major outage recently. A reactive team can be fully occupied while chronic defects, poor planning and growing backlog remain hidden. Conversely, a low maintenance spend may represent efficient asset care—or deferred work that is increasing risk.

A balanced set of maintenance and reliability KPIs helps organisations to:

• Align maintenance, operations, engineering, supply chain and finance around common priorities.
• Monitor the health of the work management process, from notification through planning, scheduling, execution and close-out.
• Identify repeat failures and bad actors that deserve reliability analysis.
• Protect capacity for preventive and predictive work rather than allowing emergency work to consume it.
• Assess whether asset availability and maintainability support the operating plan.
• Make maintenance costs visible in the context of asset criticality, risk and lifecycle value.
• Escalate emerging risks before they become production losses or major interventions.

KPIs should support professional judgement, not replace it. A falling corrective-maintenance count, for example, is encouraging only if failure reporting is complete and necessary corrective work is not being delayed or recoded.

Leading and lagging indicators: use both

A sound KPI framework combines leading and lagging indicators.

Leading indicators

Leading indicators track conditions and behaviours that influence future reliability and maintenance performance. They provide an earlier opportunity to intervene. Typical examples include planned maintenance percentage, PM compliance, schedule compliance, the age and composition of backlog, quality of work-order planning, and the completion of defect elimination actions.

These measures answer questions such as: Are we preparing work before it becomes urgent? Are preventive tasks being completed when due? Do weekly schedules reflect agreed priorities? Is critical work waiting too long for materials, access or decisions?

Lagging indicators

Lagging indicators describe outcomes that have already occurred. They include equipment availability, mean time between failures (MTBF), mean time to repair (MTTR), unplanned downtime, repeat failures, maintenance cost and, where applicable, overall equipment effectiveness (OEE).

Lagging measures show the consequences of the maintenance system, but are less useful as an early warning on their own. If availability deteriorates, the team must look back to leading indicators and failure information to determine why.

The practical balance

Avoid treating a leading measure as inherently good or a lagging measure as inherently bad. The value lies in the relationship between them. For example, improving planning and schedule compliance should eventually support more stable execution, fewer urgent jobs and better availability. If it does not, the organisation may be measuring completion rather than work quality, or it may be applying the wrong maintenance strategy to the failure mode.

Core maintenance and reliability KPIs and how to calculate them

Definitions vary somewhat between industries and CMMS configurations. The priority is to document one local definition for each KPI, state the data source and exclusions, and apply it consistently. The following are widely used general measures.

MTBF: mean time between failures

MTBF is commonly calculated as total operating time divided by the number of qualifying functional failures during the same period. It is most meaningful when applied to a clearly defined, repairable asset, duty cycle and failure definition.

Do not combine unrelated failure modes or very different operating conditions without analysis. A pump operating continuously on a demanding service cannot be compared directly with an identical spare-duty pump that runs intermittently. MTBF is also unreliable when failure records are incomplete, when operating hours are estimated poorly, or when a small number of events creates a misleading average.

Use MTBF alongside a failure-mode review. A higher value is not sufficient evidence of improvement if failures are merely being coded differently or excluded from the count.

MTTR: mean time to repair

MTTR is total active repair time divided by the number of qualifying repairs. Teams need to define what repair time includes. It may cover hands-on restoration time only, or the full elapsed outage from failure to return to service. Both views can be useful, but they must not be confused.

Long elapsed restoration can point to poor access, difficult isolation, missing spares, delayed specialist support, unclear job plans or weak troubleshooting information. Active repair time may instead highlight maintainability, tools, skills and repair methods. Separating delay categories often provides more actionable insight than one aggregate MTTR figure.

Equipment availability

Availability is generally calculated as uptime divided by the total time the equipment was required or expected to operate. The denominator must be explicit. Calendar time, planned production time and demanded operating time answer different questions.

Availability is a key outcome measure for critical assets, but it should be interpreted with downtime codes, operating demand and planned shutdowns. A high value may conceal frequent short interruptions, standby equipment carrying the load, or a period when the asset was not needed.

OEE: overall equipment effectiveness

OEE combines availability, performance and quality. Availability captures time losses, performance compares actual running rate with the ideal or target rate, and quality measures good output relative to total output. It is especially relevant for manufacturing and packaging lines where all three losses can be measured meaningfully.

OEE should not be applied mechanically to every asset. For continuous-process facilities, utilities or safety-critical equipment, availability, capacity loss, quality specifications and equipment-specific reliability measures may be more useful. Define the planned production time, ideal rate and quality criteria before using OEE for management decisions.

Planned maintenance percentage

Planned maintenance percentage compares labour hours charged to planned work with total maintenance labour hours. It is often used to assess whether the department is moving away from breakdown-driven execution.

The measure depends on clear work-order classifications. A job is not truly planned simply because it was placed on a schedule. It should have a defined scope, estimated labour, required materials, permits or access requirements, and a workable method. Review both the percentage and the quality of plans.

PM compliance

PM compliance measures how many preventive-maintenance tasks were completed within the agreed due date or allowed tolerance window, divided by tasks due in that period. It helps reveal whether the organisation is protecting routine asset-care work.

Completion should mean that the task was actually performed and properly recorded, not simply administratively closed. Critical PM tasks merit separate visibility, particularly where they address protective functions, statutory inspection requirements or known failure risks.

Schedule compliance

Schedule compliance is the proportion of scheduled work completed within the scheduled period, compared with all work scheduled for that period. It evaluates adherence to the committed work programme, not technician utilisation.

Low schedule compliance may arise because urgent work constantly displaces planned jobs, job plans are immature, materials are unavailable, equipment is not released by operations, or the schedule contains more work than available capacity. Examine causes before demanding a higher number.

Maintenance backlog

Maintenance backlog is the approved work still awaiting execution. It is commonly reported as estimated labour hours and, where helpful, converted into weeks of work using realistic available labour capacity. Backlog should be segmented by priority, asset criticality, work type, age, craft and material status.

A single total can hide serious issues. A manageable volume of future work is different from a growing quantity of overdue safety-critical, production-critical or integrity-related tasks. Track backlog ageing and investigate why work is waiting.

Maintenance cost as a share of asset value

This measure compares maintenance expenditure over a defined period with the replacement asset value or another documented asset-value basis. It provides a broad financial perspective, not a standalone verdict on efficiency.

Cost data should distinguish labour, materials, external services, emergency premiums and major improvement work where the accounting approach permits. A rising ratio may result from ageing assets, poor reliability, expanded scope, inflation, regulatory obligations or an intentional reliability recovery programme. Interpret it alongside risk, production impact and condition information.

Choosing the right KPI set

There is no universal dashboard that suits every organisation. Begin with business objectives and asset risk rather than a standard list of measures. A facility focused on stable production may prioritise availability, chronic downtime, schedule compliance and critical backlog. An ageing asset base may need stronger attention to condition findings, defect elimination and lifecycle cost. A manufacturing line may benefit from OEE, while a utility may need service reliability and capacity measures.

A practical selection process is to:

1. Identify the assets, systems and failure consequences that matter most.
2. Define the decisions each KPI should support and the person accountable for acting on it.
3. Choose a small balanced group of outcome, process and risk measures.
4. Set clear definitions, units, time periods, filters and ownership.
5. Test whether data can be captured consistently without creating unnecessary administrative burden.
6. Review the KPI set periodically as asset priorities and operating conditions change.

A hierarchy is often helpful: plant-level measures for leadership, area-level measures for weekly coordination, and asset-level indicators for reliability engineering. Do not force every team to manage the same detailed metrics.

Reliable CMMS data: the foundation of credible KPIs

A CMMS can provide a valuable evidence base, but only when the master data and work-order process are disciplined. A dashboard cannot correct missing failure codes, ambiguous asset hierarchies or hours recorded against the wrong order.

Key controls include:

• A clear asset hierarchy and consistent equipment naming.
• Defined work types, priorities, statuses and downtime categories.
• A practical failure-reporting structure, including problem, cause and remedy where appropriate.
• Accurate timestamps for notification, planning, scheduling, start, finish and return to service.
• Separate fields or codes for planned, unplanned, preventive, predictive and corrective work.
• Labour and material recording practices that match the intended cost analysis.
• Work-order close-out checks that confirm scope, findings, failure details and follow-up actions.

Data quality should be reviewed as an operational discipline. For example, managers can sample completed work orders, investigate uncoded downtime, and track overdue orders that remain technically open. Where operating-hour data drives MTBF, it should come from credible runtime records rather than assumptions.

Building dashboards and reviewing results with operations

Dashboards should make exceptions, trends and decisions visible quickly. They should not become a display of every available CMMS field. Use a consistent view of the reporting period, filter logic and asset population, then show trends rather than isolated monthly values wherever possible.

An effective dashboard may include:

• A small executive view of availability, unplanned downtime, maintenance cost and critical risk exposure.
• A work-management view of planned maintenance percentage, PM compliance, schedule compliance and backlog ageing.
• A reliability view of repeat failures, bad actors, MTBF, MTTR and open corrective actions.
• Drill-down capability from a headline measure to affected assets, work orders and downtime events.

Maintenance cannot improve availability alone. Joint operations-maintenance reviews should ask: Which losses mattered most? What work was displaced and why? Which equipment was unavailable because of maintenance, operations or external constraints? What action is required before the next review?

A weekly review can focus on schedule execution, work release, emerging urgent work and critical backlog. A monthly review can examine trends, repeat failures, cost drivers and completed improvement actions. The meeting should finish with named owners, due dates and a record of whether previous actions were effective.

Teams seeking a more structured approach to visual reporting may benefit from Maintenance & Reliability Dashboards with Excel & Power BI training, particularly when they need to translate CMMS extracts into decision-ready views.

Common mistakes in measuring maintenance performance

Several practices weaken even well-designed maintenance and reliability KPIs:

• Too many KPIs: Large scorecards dilute attention and encourage reporting rather than action.
• Unclear definitions: Different departments calculating “availability” or “completed PM” differently will produce avoidable disputes.
• Managing the number rather than the process: Closing work orders prematurely can improve compliance figures while reducing data integrity.
• Ignoring data quality: A precise-looking dashboard is not credible when failure, time and cost records are incomplete.
• Using averages without distribution: An average MTTR can conceal a small number of exceptionally long outages.
• Comparing unlike assets: Different duties, ages, process conditions and redundancy arrangements require context.
• Treating all backlog as bad: The priority, risk, age and readiness of backlog matter more than its total alone.
• Reviewing results without actions: Repeated discussion of the same trend erodes confidence if root causes and owners are absent.
• Setting arbitrary targets: Internal targets should reflect asset criticality, operating requirements, risk tolerance and data maturity—not an unexamined external figure.

Turning KPI findings into improvement actions

The purpose of measurement is controlled learning. When a KPI indicates adverse performance, move from the metric to the underlying mechanism.

For a repeat failure, confirm the event data, define the functional failure and examine the failure mode, operating context, maintenance history, spares and previous interventions. Select an action that addresses the cause: revise a PM task, improve precision maintenance, change operating practice, remove a design weakness, strengthen condition monitoring or update the job plan. Then verify whether the action changed the intended measure over time.

For low schedule compliance, separate controllable causes from symptoms. If material shortages drive disruption, improve bill-of-material accuracy and procurement lead-time visibility. If work release is unreliable, establish a clearer operations-maintenance coordination routine. If urgent jobs dominate, investigate whether preventive tasks, inspection routes or defect prioritisation are inadequate.

Reliability-centred maintenance principles can help teams match maintenance tasks to failure consequences and failure behaviour, rather than simply increasing PM volume. For organisations formalising this process, Reliability Centered Maintenance training provides a relevant learning route. Practical work-control capability can also be strengthened through Maintenance Planning & Scheduling training.

Key takeaways

Maintenance and reliability KPIs are most useful when they create a shared, evidence-based discussion about asset risk and work quality. Balance leading indicators—such as PM and schedule compliance—with lagging outcomes such as availability, MTBF and MTTR. Define each measure carefully, protect CMMS data quality, segment results by criticality and cause, and review trends jointly with operations.

Most importantly, connect each material KPI deviation to a clear investigation, an owned action and a check on effectiveness. That is how a dashboard becomes an asset-performance management tool rather than a reporting exercise.