Choosing the right preventative maintenance KPIs starts with a simple test: will the metric predict a failure soon enough to act and will it guide a clear response.
To really prevent maintenance (PM), we believe the first tier you should focus on is a short list that pairs condition with consequence so your team sees the why behind the number. Mean time between failures shows whether you are stretching asset life, but it can hide pain if you do not record every stop. Mean time to repair reveals how fast you recover when a line goes down.
Planned maintenance percentage shows whether you are doing the work on your terms instead of firefighting on the supplier’s schedule. Pair those with basic capacity measures like availability and queue time so you see how downtime affects delivery.
Your second tier should tie cause to cost. Track maintenance cost as a percent of replacement asset value so you know when to repair vs replace. Track unplanned downtime hours by workcenter and failure mode so you hunt the right issues first. Record first pass quality by machine group so you spot when wear starts to push defects. Keep the list short so people remember it.
If a KPI does not trigger a decision, drop it. Ground your picks in credible guidance and shop reality. For a structured way to think about maintenance economics, skim NIST’s analysis of advanced maintenance costs at NIST AMS 100-18. These resources help you avoid vanity metrics and select measures that predict failures and protect throughput.
A KPI only helps when it lives where the work happens. That is why maintenance data belongs in your ERP, not spread across spreadsheets.
When workorders, materials, schedules and quality records live in one system, PM metrics connect directly to delivery promises and costs. Start by assigning assets to workcenters so downtime shows up on the same calendar that production uses. Log PM tasks as routings with estimated hours, then record actuals with barcode transactions. Tie spare parts to items in inventory with min and max levels so PMs do not stall waiting on a bearing, for example.
Attach procedures, photos and checklists to each PM using ERP Document Control so techs know the standard and supervisors can audit quickly. Make it easy to capture the facts. Place rugged devices and scanners at the maintenance bench and in problem cells. Keep any visible monitors angled away from aisles and show only the fields needed to clock on, consume a part or close a task. Use simple failure codes so the history is searchable.
With these basics in place, your dashboards can display planned vs unplanned hours, PM completion rate by asset and parts consumption by class. That is the visibility leaders need to protect on-time delivery. If you want a quick primer on the tools that support this flow, review our Preventative Maintenance application and the broader ERP suite on our ERP Software Applications page. These pages outline how maintenance ties to scheduling, inventory and purchasing so PMs never become a paperwork exercise.
Uptime grows when PM becomes a daily habit, not an annual goal. Set a short drumbeat of reviews that keep eyes on the right numbers. In each cell, post yesterday’s planned vs unplanned maintenance hours and the top three failure codes. In the weekly ops meeting, review MTTR on the five assets that drive most late jobs and agree on one countermeasure per asset. In the monthly review, check maintenance cost as a percent of replacement asset value so you stay honest about repair vs replace decisions. Link each review to a named owner and a due date so actions do not drift.
Use your ERP to make coaching easy. Build dashboards that show trends by workcenter and drill into the job, the part and the operator notes without leaving the screen. When PM completion drops, trigger a message to maintenance and the area supervisor. When a spare hits min, auto-create a PO suggestion tied to the next PM date. Connect quality to maintenance by reviewing scrap and rework by machine family in the same meeting.
These are simple moves that turn metrics into throughput. For more background on how manufacturing organizations frame KPIs and maintenance strategy, see NIST’s overview on enhancing maintenance at enhancing maintenance strategies. And for an example of how AI is improving maintenance insight inside ERP, this piece shows practical gains at AI in ERP 2026. Keep the list short, the capture simple and the cadence steady, and your PM program will lift uptime without adding drag.