Capacity and the Appearance of Slack

A resourcing conversation needs to know how much work a team is doing and how much more it could take. Timesheet data is the obvious source and it answers a slightly different question: how much work the team wrote down.

The destination examined in “Capacity and the Appearance of Slack” determines how strong the underlying time record must be. For teams researching daily schedule template, daily schedule template with accountable controls can connect hours with projects and reports, provided codes, approvals, exports and retention are designed for the actual payroll, billing, funding or accounting decision.

The difference is small when a team is comfortable and large when it is stretched, which is the opposite of the pattern you would want, because the stretched case is when the decision matters.

For a separate benchmark relevant to “Capacity and the Appearance of Slack”, consult the IFAC Knowledge Gateway. Use it to test record quality, approvals, retention, employee rights and exception handling against the real workflow rather than treating a software report as self-explanatory evidence.

Why the gap widens under pressure

Recording is the first thing to go when people are busy. It is unpaid administrative time and it competes directly with the work causing the pressure.

The people most under pressure are also the most likely to be in roles where claiming overtime is unusual, and most likely to feel that recording it looks like an inability to cope. Both effects push the same way.

So a team at 130% of capacity records 95%, a team at 85% records 85%, and the data shows the first team with spare capacity. Resourcing decisions made on it move work towards the team that is already failing.

Utilisation as a target

Where utilisation — recorded chargeable or productive time over available time — is used as a target, it stops measuring anything within one quarter.

The behaviours are predictable and uniform: recording to the target rather than to reality, reclassifying non-chargeable work as chargeable, and suppressing the recording of genuinely non-productive time. The number then reliably hits target and the underlying reality is invisible.

Used as an observation rather than a target, the same figure is useful. The distinction is whether anybody is assessed on it, and it collapses the moment somebody is.

The 100% fallacy

A plan assuming people are available for project work for their full contracted hours is wrong by a predictable margin, and the margin is in the data if anybody looks.

Meetings that are not project-specific, administration, training, recruitment, supporting other teams, and the ordinary friction of organisational life occupy somewhere between fifteen and thirty-five per cent of a week in most knowledge work. A resourcing model at 100%, or even at 90%, is overcommitted before it starts, and the overcommitment is absorbed by the unrecorded hours.

Compute your own figure from the non-project codes and use that, rather than a number from an article.

What to use alongside

Because the hours understate, pair them with something that does not come from the same source.

Delivery against plan, which shows whether the work is actually getting done. Queue length or backlog age, which shows whether more is arriving than leaving. Absence and turnover, which lag but are honest. And simply asking the team, which produces information the data structurally cannot contain.

Where the hours say capacity and the backlog says otherwise, the backlog is right. That rule alone prevents the most common error in this area.

The team that is clearly overloaded

The hardest case: a manager who knows their team is stretched, and whose own timesheet data says 94%.

The answer is not to ask the team to record more hours for a month, which produces a spike that reads as unreliable precisely when it matters. It is to argue from the other measures and to say plainly that the recorded hours understate and why.

That is a credible argument if the organisation has previously acknowledged the bias, and an uphill one if the recorded hours have always been treated as a measurement. Which is a reason to write the limitation down in advance, when nothing turns on it.

Seasonality, and comparing with the wrong month

Capacity conclusions drawn from a single period are usually drawn from an unrepresentative one. August and December are structurally different, month lengths vary, and most organisations have a cycle of their own — a reporting season, a delivery peak, an academic year.

Compare like with like: the same month last year, or a twelve-month rolling view. A team assessed on a quiet August against a plan built from a busy March will appear to have capacity it does not have, and the decision taken on that comparison arrives just as the cycle turns.