Skip to content
What the Plan Already Owes

Home / Variance

A Plan You Can Defend

What makes a capacity plan survive questioning: the four properties, what to do about the figures you cannot evidence, and how a defensible plan behaves when it turns out to be wrong.

Variance · Reference

What a defensible plan can produce when each figure is questioned

Net capacity3,180 hoursConfirmed

Measured supply less dated shrinkage.

Committed before this plan2,460 hoursConfirmed

Signed off, with names and dates.

Standing support520 hoursProvisional

Re-measured in February; next check in August.

Uncommitted estimate310 hoursAssumed

Pipeline, carried at an estimating rate.

Remaining-110 hoursConfirmed

The shortfall, arithmetic, before any choice is made.

Three of the five lines carry a source and a date; the two that do not are labelled. A plan is defensible when the labels are visible, not when every line is firm.

A defensible plan is not one whose numbers turn out to be right. It is one where every figure can be traced to a source, a date, or an explicit assumption, and where the assumptions are labelled as such.

The reconciliation in “A Plan You Can Defend” is strongest when assumptions and current work records can be examined separately. A team considering review the complete feature explanation for employee monitoring at tech companies can add time and project context, with transparent settings, limited access and a clear correction route around consequential decisions.

That is the whole of it, and it is the reason the account described in this collection has a status column.

The assumptions behind “A Plan You Can Defend” can also be tested against APQC process frameworks. Record which definitions transfer to the local plan and which do not.

The four properties

Every figure has a source. Every figure has a date. Every commitment has a name against it. And the difference between what exists and what is promised is stated rather than closed.

A plan with those four properties survives being questioned by someone who did not build it, which is the only test that matters.

The figures you cannot evidence

There are always some: a pipeline item, a service level nobody has measured, next year's attrition.

They go in labelled as assumed, with whatever basis exists. An unevidenced figure presented as firm is the single defect that makes a whole plan unusable, because it means the labels cannot be trusted anywhere.

Why the shortfall stays visible

Because a plan that balances on paper either contains a committed figure that was quietly reduced or an assumed figure that was quietly raised.

Leaving the negative number on the page is what makes the rest of the arithmetic believable, and it is what moves the choice to the people who are allowed to make it.

Being wrong well

A defensible plan that turns out wrong produces a correction to a named line with a date. An indefensible one produces an argument about whether anybody could have known.

The first improves next quarter's plan. The second produces a commitment to communicate better, which is not a figure and cannot be checked.

What the questioner is entitled to

Any single figure's source, on request, without preparation. If the answer is "I'd have to go and look", the account is not finished; if it is "that's assumed, on this basis", the account is doing its job.

A plan that needs its author present to be understood is a presentation, not an account.

What it is not

It is not a forecast, and it does not claim to predict the quarter. It states what exists, what has been promised, and the gap, each as of a date.

Capacity planning that promises accuracy sets itself an impossible test and fails it publicly. Capacity planning that promises traceability passes a possible test every time.

The maintenance

A close each period: claims closed, variance recorded, a line or two corrected. A recalculation of shrinkage once a year. A re-measurement of support every second year.

A few hours a quarter. The alternative is not less work — it is the same work done as an argument, at the point where the shortfall has already landed on somebody.

What changes in the organisation

The commitments get smaller and more specific, because a commitment that has to carry a name and a size is harder to make casually.

That is the main effect of running a visible account, and it happens without anybody being told to commit to less.

Where to start, if nothing exists

With one team, one quarter, four lines: measured supply, dated shrinkage, signed commitments, and the difference. A page.

That page, honest about what it does not know, is more defensible than a complete model built on figures nobody can trace, and it takes an afternoon rather than a project.

The plan that was never questioned

Worth noticing, because it usually means the plan is not being used. An account that nobody interrogates is an account nobody is relying on, and the absence of questions is not a sign of quality.

The first quarter of hard questions is the sign that the document has started to do something.

Handing it over

A defensible plan can change owner without a handover meeting, because every figure carries its own source and date.

That is a practical test anybody can run on their own account: if it would take a week of explaining to pass on, the explanations belong in the document.

The last thing

The number at the bottom of a capacity plan is not the output. The output is the set of labelled rows above it, because that is what anybody can act on: a stale figure to re-measure, a commitment to question, an assumption to confirm.

A plan read for its bottom line will be argued with. A plan read for its rows gets corrected, which is the only way it ever becomes right.