Maintenance Performance Metrics & KPI Reporting

Reliability Operating Centre

Maintenance Performance Metrics & KPIs

Measures your team can act on, not just report. A metric set with agreed definitions, a clear line from daily maintenance activity to business outcome, and an owner on every number.

Three ways a metric set stops being useful

Most operations are not short of numbers. They are short of numbers anyone acts on.

Definitions have drifted

Two reports, three definitions of downtime, and the first twenty minutes of the meeting spent arguing about whose number is right instead of what to do about it. Once that happens twice, people stop bringing data.

Everything is lagging

Availability and MTBF tell you last month went badly. They do not tell anyone what to change on Monday. Without leading measures there is no way to know you are off course until the result has already happened.

There are too many

Forty KPIs is not a measurement system, it is a spreadsheet. When everything is measured, nothing is prioritised, and the reporting effort quietly exceeds the value of the reporting.

If a metric has no owner and no action attached to it, it is not a metric. It is a number.

What we do

Build the value treeStart at the business outcome, whether that is tonnes, cost per unit or availability, and work down to the daily activities that move it. Every metric has to earn its place by sitting on a branch of that tree — the dollar branch is where loss accounting takes over. Anything that does not connect comes off the report.

Write the definitions dictionaryOne agreed definition per measure, with the calculation, the data source, the boundary conditions and the exclusions written down. Dull work, and it is the single highest-return thing most sites can do to their reporting.

Balance leading against laggingPM compliance, schedule compliance, strategy currency and work order data quality tell you whether next quarter will be better. Availability, MTBF, MTTR and unit cost tell you whether last quarter was. You need both, and most sites only report the second kind.

Automate the collectionPulled from your CMMS on a repeatable extract rather than rebuilt by hand each month — the same discipline our maintenance management work builds into planning and scheduling. If producing the pack depends on one person remembering a process, the pack stops when they take leave.

Put a name against every measureMOST OFTEN SKIPPEDEach metric gets an owner who is accountable for the number and for the action when it moves the wrong way. A report where nobody owns anything is a newsletter, and it will be read like one.

What you get

Deliverable	What it contains	Cadence
Value tree	The hierarchy linking business outcomes to maintenance activity, on one page	Annual review
Definitions dictionary	Every measure with its calculation, source, boundaries and exclusions	Live, version controlled
Monthly performance pack	Leading and lagging measures, trends, exceptions and explanatory commentary	By the 5th business day
Exception alerts	Notification when a measure breaches its control limit, not at month end	As they occur
Benchmark view	Your performance against comparable equipment in our library	Quarterly

Frequently asked questions

What KPIs should a maintenance department track?

A balanced set of leading measures (PM compliance, schedule compliance, strategy currency, work order data quality) and lagging measures (availability, MTBF, MTTR, unit cost) — each connected to a business outcome through a value tree, each with a named owner. Any measure that does not connect to an outcome comes off the report.

What is the difference between leading and lagging maintenance indicators?

Lagging indicators tell you whether last quarter was good — availability, MTBF, unit cost. Leading indicators tell you whether next quarter will be — PM compliance, schedule compliance, data quality. Most sites only report the first kind, which means they learn about problems after the result has already happened.

How many maintenance KPIs are too many?

When everything is measured, nothing is prioritised. Forty KPIs is a spreadsheet, not a measurement system. The test is the value tree: every metric has to earn its place on a branch linking daily activity to a business outcome, and the reporting effort must stay smaller than the value of the reporting.

Why do KPI programmes stop being used?

Three failure modes: definitions drift until meetings argue about whose number is right; everything reported is lagging so nothing is actionable; and no measure has an owner, which turns the report into a newsletter. The fix is a definitions dictionary, a leading and lagging balance, and a name against every number.

Find out whether your numbers agree with each other

The definitions review is part of Foundation. It is usually the first time a site sees its reporting reconciled end to end.

Book a fit and data readiness check