Where the Error Actually Was
The five places a capacity plan can be wrong, how to tell which one produced a given variance, and why the answer is rarely the one everybody assumes.
When a plan and reality disagree, the error is in one of five places, and identifying which one takes about twenty minutes. Most organisations skip the identification and apply a general correction, which is how plans acquire layers of margin that fix nothing.
The variance in “Where the Error Actually Was” requires an unchanged plan and a trustworthy record of actual work. Teams exploring read more in this official guide for employee monitoring software with screenshots can strengthen the actual side of that comparison while preserving corrections, category changes and human explanations.
The five are: the supply figure, the claim estimate, the intake, the sequencing, and the plan being right while the work changed.
Additional context for “Where the Error Actually Was” is available from IPMA competence standards. Use external material to prompt a specific validation check rather than replace local judgement.
The supply figure
The hours were not there. Absence above the rate, a late start, an unplanned event, a deduction that was too small.
Test: compare planned available hours with actual available hours. If the gap is there, the error is here and the remedy is a rate rather than anything about the work.
The claim estimate
The hours were there and the work needed more of them. The classic overrun.
Test: actual hours on the claim against its estimate. Where this is the error, the remedy belongs to estimating practice and to the claimant, not to the capacity account.
The intake
Work arrived that was never in the book and consumed hours anyway.
Test: hours spent on unplanned work. A large figure here means the account is being bypassed, and no improvement to supply figures or estimates will help until the intake is fixed.
The sequencing
Total hours were adequate and they were not available when the work needed them. The quarter balanced and week six did not.
Test: the monthly or weekly shape against the claims' own windows. This error produces a plan that looks correct in retrospect and failed in practice, which is the most confusing variety to diagnose.
The plan was right and the work changed
Scope grew, the customer changed their mind, a dependency moved. Nothing about the capacity account was wrong.
This is the most common cause and the least often recorded, because recording it requires somebody to say that the work changed, which sounds like an excuse when it is simply what happened.
The order to test in
Supply first, because it is the quickest and the most mechanical. Then intake, because a large figure there invalidates everything else. Then sequencing. Then the claim estimates. Then scope.
Running it in that order resolves most variances in under half an hour and prevents the common error of adjusting estimates when the problem was intake.
Why the assumed answer is usually wrong
Most organisations assume the error is estimating, because that is the visible part: a claim overran. In practice the supply figure and the intake together account for more variance than estimates do, in most places that have measured it.
Correcting estimates when the cause is intake produces inflated estimates and an unchanged problem, which is the familiar state of plans that have been "improved" for several years.
Recording the cause
One word per variance, at the close, in the same row as the numbers. Supply, estimate, intake, sequencing, scope.
After four periods the distribution of that word is the organisation's real planning problem, stated in a form that can be acted on.
What to change, by cause
Supply errors change a rate. Estimate errors change an estimating practice. Intake errors change a process. Sequencing errors change the planning period. Scope changes change nothing in the account and belong in a different conversation.
Five causes, five different remedies, and only one of them is about anybody's estimating.
The five tests, in order
- Planned available against actual available — supply.
- Hours on unplanned work — intake.
- Monthly shape against claim windows — sequencing.
- Actual against estimate, per claim — estimating.
- Whatever is left — scope change.
Twenty minutes, five subtractions, one word recorded. It is the whole of variance analysis for a capacity account, and anything more elaborate will not be run after the second quarter.
The variance that belongs to nobody
Scope change is the most common cause and the one with no owner: the work grew, everybody acted reasonably, and the plan was overtaken.
Recording it as scope rather than as an overrun protects the estimate series from noise it should not carry, and it is also simply accurate. Teams that record every scope change as an estimating failure end up with estimates padded against a risk estimating cannot address.
The discipline this requires
Resisting the single correction. A plan that was wrong by fifteen per cent invites a fifteen per cent buffer, and the buffer will not be in the place the error was.
Twenty minutes of diagnosis saves a year of margin applied to the wrong line, and it is the whole reason to keep the variance record at all.