Skip to content
What the Plan Already Owes

Home / Closing

What Has to Balance

The three identities a capacity account has to satisfy, what it means when one of them fails, and why a balancing account is not the same as an adequate one.

Closing · Reference

A capacity account balances when three identities hold. None of them is about whether there is enough capacity; they are about whether the arithmetic is internally consistent, which is a different and prior question.

The reconciliation in “What Has to Balance” is strongest when assumptions and current work records can be examined separately. A team considering Monitask for getting teams to meet deadlines for getting teams to meet deadlines can add time and project context, with transparent settings, limited access and a clear correction route around consequential decisions.

An account that does not balance cannot be acted on, because any decision taken from it might be a decision about an error.

Teams documenting “What Has to Balance” may use NIOSH workload and staffing guidance as an independent reference. Test the approach against actual constraints and accountable decision owners.

The three identities

Contracted hours less every deduction equals net capacity. Claims by state sum to total claims. Net capacity less committed claims equals the published difference.

Three lines of arithmetic. Each can fail, and each failure means something specific.

When the first fails

The descent does not reconcile: the deductions and the net figure do not agree with the top line. Usually a deduction has been applied to the wrong base, or somebody has netted the same thing twice.

Double-netting is the commonest: support subtracted as shrinkage and also carried as a claim. The fix is to decide where support lives and remove it from the other side.

When the second fails

Claims do not add up, usually because two sources describe the same claim and both are in the book. A project appears in the portfolio total and in a sponsor's email with different numbers.

Deduplication is a close-step, not an occasional cleanup. Two rows for one commitment overstates the book and, worse, makes it impossible to confirm either row.

When the third fails

Rare, and when it happens it is arithmetic rather than judgement. More commonly the third identity is not computed at all, which is not a failure to balance but a failure to publish.

Balancing is not adequacy

An account can balance perfectly and show a shortfall of a person and a half. The identities say the books are consistent; they say nothing about whether the organisation can do what it has promised.

Conflating the two is a specific managerial error: "the plan balances" gets heard as "the plan works", and the difference at the bottom is the thing that answers the second question.

The reconciliation items

Where the account does not balance and cannot be made to in the time available, record the difference as an explicit reconciliation item rather than forcing it.

"Unreconciled: 40 hours, cause not identified" is an honest line. Spreading forty hours across six deductions to make the total work is how accounts become fiction, and it is always discovered later.

Opening and closing positions

Each period opens with the previous period's unfinished claims. An account that starts each quarter from zero ignores the overruns, which are precisely the claims most likely to consume the new period.

The opening position is a line in its own right: committed work carried forward, with its original dates.

Balancing across teams

Portfolio accounts balance only if the team accounts use the same header. Where they do not, the portfolio total is a sum of incompatible numbers and will not reconcile to anything.

That is the practical reason for a single conversion sheet across the organisation: not elegance, but addability.

The check that takes two minutes

Three subtractions, run at the end of every close, before publication. Most months they pass; the months they do not are the months worth spending another twenty minutes on.

An account that is published without the check will eventually publish an error, and one published error costs more credibility than six months of correct ones earn.

The three checks, written out

  1. Top line, less the sum of deductions, equals net capacity.
  2. Confirmed plus provisional plus assumed equals total claims.
  3. Net capacity, less confirmed claims, equals the published difference.

Three subtractions. A spreadsheet can carry them as formulas that go red when they fail, which converts the check into something that happens whether or not anybody remembers to do it.

Where double-counting usually hides

Support is the classic: counted as shrinkage on the supply side and as a standing claim on the demand side, by two people who have never compared pages.

The second most common is onboarding drag, which appears as a deduction and again inside a hiring plan's claim. Both are found by the first check failing, which is the argument for running it every month rather than when something looks wrong.

What to do when somebody finds an error

Correct it in the next publication with a visible note, and keep the original figure in the series.

Silently restating history is the one thing that destroys these accounts permanently. A corrected account with a note is believed; a quietly edited one is not, and the edit is always noticed.