Staffing to Actual Demand

A team that functions only when everyone is present is not adequately staffed. Absence is normal, not exceptional.

The arithmetic nobody does

Expected availability is not headcount. Subtract leave, public holidays, sickness at your historical rate, training and administration.

In many organisations, planned availability is around seventy to eighty percent of headcount across a year, and planning assumes a hundred.

The gap is absorbed by overtime, deferred work, or the queue.

Then a vacancy opens and the gap becomes structural.

Why lean staffing is expensive

Overtime costs more per hour where it is paid, and costs attrition where it is not.

Errors rise with fatigue, and in some sectors that is a safety and regulatory cost rather than a productivity one.

Nobody has time for improvement, so the inefficiencies generating the load persist.

Absence becomes catastrophic, so people work while ill, extending illness and spreading it.

Departures compound. Each one increases the load on those remaining, raising the probability of the next.

Recruitment gets harder as the reputation spreads, extending vacancy duration, which deepens the problem.

Coverage that actually works

Cross-training, deliberately, so that no function has a single point of failure. This has a cost in time and it is the cheapest form of resilience available.

Documented handover for anything one person holds alone.

Named cover for leave, agreed before the leave rather than during it.

A genuine buffer. Planning at a fraction below full capacity rather than at it.

Prompt backfill. Vacancy duration is a wellbeing metric and should be tracked as one.

The single point of failure

Worth identifying explicitly.

List the functions only one person can perform.

For each, ask what happens if they are away for two weeks unexpectedly.

The list is usually longer than expected and includes people nobody thinks of as critical.

Each item is both an operational risk and a wellbeing problem, because that person cannot take leave properly and knows it.

Making the case

Use the arithmetic, not the sentiment. Demonstrated throughput against committed volume, with the availability calculation shown.

Use turnover cost from your own data, per departure, against the cost of the additional person.

Use vacancy duration and its trend.

Show what is being deferred, which is frequently invisible to whoever approves headcount because the team has quietly protected them from it.

When the answer is no

Sometimes headcount is genuinely unavailable.

Then something must be dropped, and the decision about what should be made by whoever declined the headcount, not left to the team.

Present the list of what will not be done and ask for a decision on priority. This is not a confrontation; it is the only honest response to a fixed capacity and a growing demand.

Record the decision. Teams end up in permanent overload because this decision is repeatedly avoided rather than because anyone chose it.

The single-point-of-failure list

An hour of work that produces both an operational risk register and a wellbeing finding.

List every function only one person can perform.

For each, note what happens if they are unexpectedly away for two weeks.

Note whether that person can currently take leave properly. Usually they cannot, and they know it.

Rank by consequence.

Cross-train against the top of the list, allocating actual time rather than hoping it happens.

Document what cannot be cross-trained quickly.

The list is usually longer than expected and includes people nobody thinks of as critical — which is precisely why their absence is unplanned for.

Vacancy duration as a wellbeing metric

A number already sitting in a system, rarely read this way.

Days from departure to replacement starting, by team.

Every day is load absorbed by the remaining people.

Track the trend, since lengthening vacancies compound.

Read it alongside turnover. A team with rising turnover and lengthening vacancies is in a loop that will not resolve on its own.

Report it to whoever approves headcount, because it converts an abstract staffing request into a measured accumulating cost.