The short version: schedule compliance measures how much of the work you committed to at schedule freeze was finished inside that schedule period. Most operations measure something loosely related and give it the same name. Six habits inflate it: editing the frozen schedule after the fact, counting jobs instead of hours, reading completion off the order status instead of the confirmations, leaving breakdown work out of the denominator, rolling deferred work forward as fresh, and never agreeing which day the schedule actually freezes. Fix the definition before you chase the number. A wrong number still gets reported upward, and once it is reported it gets defended.
I have spent nine years inside work management on Anglo American operations, across four commodities. The complaint I hear most often is rarely that compliance is low. It is that compliance reads eighty five percent and the week still felt like a fire.
Both things are usually true at the same time. That is the problem worth writing about.
What is schedule compliance actually measuring?
It measures a promise. At schedule freeze you committed to a set of work for a period. Compliance asks how much of that committed work was finished inside the period.
The word doing the work in that sentence is committed. Everything else the crew did that week, however hard they worked and however necessary it was, sits outside the measure by design. That is the point of it. Schedule compliance reports planning integrity, and the hours worked belong to a different measure.
Once you hold that definition steady, the six traps below turn out to be versions of one mistake: changing what counts as committed.
Why does compliance stay high while the week falls apart?
The denominator moves
Compliance gets calculated against the schedule as it stood on Friday afternoon, after a week of edits. Work pulled out on Wednesday never counts as missed, because by Friday it was no longer in there.
Freeze the denominator at schedule freeze and store it. If your system lets somebody edit a frozen schedule without leaving a record, your compliance number is describing a document that has been rewritten to agree with events.
Jobs get counted instead of hours
A thirty minute lubrication route and a forty hour gearbox change each count as one job. A crew can reach eighty five percent by closing every small job and deferring the one that mattered.
Measure in scheduled hours. The arithmetic changes immediately, and so does the conversation in the Monday meeting.
Completion gets read off the wrong field
An order sitting at TECO with operations only partly confirmed is closed paperwork over open work. TECO is a system status and it tells you what an administrator did. The operation confirmations tell you what an artisan did.
Read compliance off the confirmations. In SAP PM a fully confirmed order carries CNF and a partly confirmed one carries PCNF, and the gap between those two is the gap between finished and filed.
Breakdown work sits outside the denominator
Unplanned work gets excluded on the grounds that it was never scheduled. That is arithmetically defensible and operationally dishonest, and it produces the situation every planner recognises: the compliance line climbing while the schedule is destroyed daily.
Report schedule compliance and schedule break in the same report, every week, side by side. One number without the other is half a sentence.
Deferred work reappears as new work
A job deferred from last week turns up in this week’s schedule and counts as fresh. Nothing in the percentage records that it has now moved four times.
Age the deferred work. A job deferred four times is a planning failure, and no compliance percentage will ever show it to you.
Nobody agrees when the schedule freezes
Ask three people in the same department which day the schedule freezes and what may still enter it afterwards. Three answers means the meaning of scheduled is drifting week to week, and your trend line is measuring that drift rather than your performance.
How do you calculate it correctly in SAP PM?
Four decisions, made once and written down.
- Store the frozen schedule. Snapshot the committed operations and their scheduled hours at freeze, then keep the snapshot. Everything afterwards is measured against that stored set.
- Count scheduled hours rather than orders. Numerator and denominator both in hours.
- Take completion from the operation confirmations, inside the schedule week. A confirmation captured late lands in the wrong week and shows up as work completed late, which is a different fault with a different cure.
- Separate the work types with user status. Urgent work carries URG. Without that status you cannot pull unscheduled work out of the scheduled population, and the measure collapses into an average of two unlike things.
That fourth point gets skipped most often, and it is the cheapest of the four to fix.
What should you look at instead of one percentage?
A single compliance figure cannot carry a week. Put four numbers next to each other and the picture assembles itself.
| Measure | What it exposes |
|---|---|
| Schedule compliance, in hours | Whether the plan survived the week |
| Schedule break, in hours | What displaced it, and who authorised that |
| Deferral age | Work the schedule keeps pushing rather than resolving |
| Confirmation timeliness | Whether the data behind the other three can be trusted |
Confirmation timeliness belongs on that list for a reason that has little to do with administration. If confirmations arrive a week late, every other number on the page is describing a week that has not finished reporting itself.
What does a good number look like in a South African operation?
The benchmark usually quoted is above eighty percent, and it came out of American reliability practice. It travelled here without its definition, which is the part that matters.
An operation that measures in hours, against a stored frozen schedule, with breakdown work reported separately, often finds its real compliance sits in the fifties or sixties the first time it looks. The operation has not got worse. It has become visible.
The useful target is your own trend, measured the same way for twelve consecutive weeks. A number you can compare against yourself beats a number you borrowed from a plant in Ohio.
DBiLL builds work management capability in operations running SAP PM, from the measures through to the roles that feed them. If your compliance number and your week disagree, book a free delivery review and we will look at how it is being calculated.
Related reading: the five phases of work management.
Frequently asked questions
What is schedule compliance in maintenance?
It is the proportion of work committed to at schedule freeze that was completed within that schedule period, measured in scheduled hours. It reports whether the plan held, rather than how much work the team did.
Why is my schedule compliance high when everything feels reactive?
Almost always because breakdown and urgent work has been left out of the denominator, or because the frozen schedule is being edited during the week. Either one lets the percentage rise while the schedule is overwritten daily. Report schedule break alongside compliance and the contradiction resolves itself.
Should schedule compliance be measured in hours or in job count?
Hours. Counting jobs treats a thirty minute lubrication route and a forty hour gearbox change as equal, which lets a crew score well by completing small work and deferring the job that mattered.
What is the difference between TECO and CNF in SAP PM?
TECO is technically complete, a status set on the order to close it off and release its reservations and requisitions. CNF is set once every operation on the order has been confirmed. An order can sit at TECO while operations remain unconfirmed, which is why compliance should be read from confirmations rather than from the order status.
When should a confirmation be captured?
Inside the schedule week the task was scheduled in. Compliance measures run on week buckets, so a confirmation captured late lands in the wrong bucket and misreports two weeks at once.
What is a realistic schedule compliance target?
Set it against your own baseline rather than an imported figure. Measure the same way for twelve weeks, then set the target off that trend. Operations that start measuring this way usually find their true starting point sits well below the eighty percent they had been reporting.
Does urgent work count against schedule compliance?
Only if it displaced committed work, and you can see that only if urgent work carries the URG user status. Without that status, unscheduled work cannot be separated from scheduled work and the measure stops meaning anything.