Why Planned Maintenance Fails — and What To Do About It

The plan is usually fine. The execution is not.

The first thing we do when a site asks us to look at their maintenance performance is pull the PM compliance data. Nine times out of ten, the schedule itself is broadly reasonable. The problem is that somewhere between the work order being raised and the job getting done, things fall apart.

We have seen sites with 40% PM compliance telling themselves they have a maintenance strategy. They do not. They have a list of things they intended to do. That is a very different thing. And when 60% of your planned work is slipping, you are not preventing failures — you are just documenting them slightly earlier.

The reactive cycle is self-reinforcing. Planned work gets bumped because there is a breakdown to deal with. The breakdown happened partly because the planned work was not done. Nobody has time to notice the pattern because there is always another fire to fight.

What actually causes it — from 15 years on the floor

Production takes the maintenance window and never gives it back

This is the most common one and the hardest to fix because it involves a cultural battle that maintenance teams rarely win head-on. A planned window gets offered to production and production says they need it. The PM gets deferred. Then deferred again. By the time the fourth deferral comes round, nobody is even logging it properly — it just disappears into the backlog.

The fix is not trying to outrank production in the pecking order. It is making the cost of deferral visible. We have sat in planning meetings and said "that pump has now missed its bearing check three times — here is what the last two failures on that asset cost you in downtime." That changes the conversation. Finance get interested. The decision to defer suddenly has a number attached to it.

Nobody actually owns the schedule

Ask a maintenance team who is responsible for PM compliance and you will often get a shrug or three different answers. The planner thinks the supervisor owns it. The supervisor thinks the planner manages it. The maintenance manager assumes it is all ticking along fine until the monthly report lands and it clearly is not.

Compliance needs a named owner. Someone whose job it is to track it week by week, understand why tasks are slipping and escalate when patterns appear. That does not require a new hire — it requires making it explicitly somebody's accountability rather than nobody's.

The task times are wrong and everyone knows it

Many PM schedules were written years ago and have never been validated against what the job actually takes. When a technician is allocated two hours for a job that takes four, one of two things happens. Either they rush it and do it badly, or they do half of it and close it off. Both outcomes look like compliance on paper. Neither actually protects the asset.

We make a habit of walking PM tasks with technicians before recommending any changes to frequencies or strategies. You learn a lot standing next to someone trying to do a bearing inspection on a motor that is three feet off the ground with no permanent access platform. The job plan says 45 minutes. It takes 45 minutes just to get safe access sorted.

The CMMS has been poisoned by bad data

If the system has wrong equipment records, tasks that generate for assets that were scrapped two years ago, or job plans that nobody has touched since they were imported from a spreadsheet in 2019 — people stop trusting it. They start keeping their own lists. The planner has a spreadsheet. The supervisor has a whiteboard. The CMMS becomes a reporting tool for management rather than a planning tool for the team that is actually doing the work.

We have seen CMMS systems that are technically world-class and practically useless because of data quality. A focused clean-up of the 20% of records that cause 80% of the confusion will do more for maintenance performance than any new module or upgrade.

There is no feedback loop

On a well-run maintenance system, a completed PM task generates information. The technician found early bearing wear on Asset 14. Running temperatures on the compressor are creeping up. The gearbox oil smells burnt. That information should feed into the next planning cycle — tighten up the frequency, flag for closer monitoring, bring the oil change forward.

What actually happens on most sites is that the task gets signed off and closed. The findings live in the head of the technician who did the job. No one looks at them. The schedule runs on the same assumptions it always did, regardless of what the equipment is telling you.

Where to start if your site is stuck in this cycle

Do not start by redesigning your maintenance strategy. Start by understanding what is actually happening with the one you have got.

Four questions worth answering honestly before anything else:

• What is your actual PM compliance rate — and has anyone measured it in the last 90 days?
• Which assets have the most overdue tasks, and which have the most failures? Is there overlap?
• What are the most common reasons jobs get deferred or closed incomplete?
• Does your maintenance team trust the data in your CMMS enough to plan from it?

Those four answers will tell you more about where to focus than a full strategy review. In our experience the fixes that follow are usually practical and unglamorous — sorting out ownership of the schedule, validating task times on the ten most critical assets, running a data clean-up on the parts of the CMMS that are causing the most friction. None of it is exciting. All of it works.

If you do not have the bandwidth to step back and look at your system while you are running it — which is exactly the problem, and exactly why it perpetuates — that is where we come in.