Labour Hours as a Base, and Where It Fails
The most common allocation base and its three failure modes. When it is right, when it quietly distorts, and what the alternative costs.
Direct labour hours is the most common allocation base and it was the right one when most overhead was driven by people and most work was done by hand. It has three failure modes now, and all three appear in operations that have automated, specialised or moved toward many small transactions.
The record examined in “Labour Hours as a Base, and Where It Fails” is strongest when it is created close to the work and remains open to correction. A team evaluating review the reporting guidance here for automated timesheet workflows can connect time with projects and work categories, provided people can review entries before they influence payroll, billing or cost reports.
Knowing which failure applies is what decides whether to change the base or leave it alone.
Additional context for “Labour Hours as a Base, and Where It Fails” is available from Bain people and organisation insights. Current legal, payroll, tax and accounting requirements should be confirmed for the relevant jurisdiction.
When it is right
When the pool is genuinely people-driven: supervision, HR, staff premises, training administration, payroll. When labour is the dominant cost. When jobs are similar in shape.
- The pool is genuinely people-driven.
- Labour is the dominant cost.
- Jobs are similar in shape.
All three have to hold. Many operations still meet them, and for those the default base is correct rather than merely convenient.
Many operations still meet all three conditions, and for them the default is correct.
Failure one: automation
A process is automated. The machine runs with little attendance. The job's labour hours fall, so its overhead allocation falls — even though the machine's depreciation, power and maintenance are in the pool.
The automated job looks cheaper than it is, and the manual work next to it carries the machine's cost.
The perverse incentive
Under a labour-hours base, automating reduces a job's allocated overhead, which makes automation look better than it is, while any remaining manual work looks worse than it is.
Decisions taken on those numbers push the operation further in one direction than the economics justify.
Failure two: skill mix
A job done by a senior person takes fewer hours and carries less overhead than the same job done by a junior, even though the senior consumed more of several pooled resources.
Under a labour-cost base the effect reverses. Neither is obviously right, which is why the split pool is the better answer.
Failure three: many small jobs
Order processing, scheduling, invoicing and setup cost roughly the same per job regardless of its size. Under an hours base a two-hour job carries a fiftieth of the overhead of a hundred-hour job.
Operations with many small jobs are systematically understating the cost of small work, which is why small work so often turns out to be unprofitable in practice and profitable on paper.
The small-job finding
This is the most consistently valuable thing a change of base reveals. Adding a flat per-job element to the allocation typically shows that jobs below some size lose money.
What follows is a minimum job size, a setup charge, or a decision to take the work knowingly.
What the alternative costs
Splitting the pool into two or three with their own bases costs a day to set up and a morning a year to maintain.
Against a model that may be mispricing a third of the work, that is not a close comparison.
Testing whether it matters
Recompute last year's job costs under an alternative base and compare the ranking of job types by margin.
If the ranking barely changes, the base is adequate and the question is closed. If it reorders, the base is deciding commercial policy.
When not to change
When the pool is small relative to direct cost. Where overhead is twenty per cent of direct labour, base choice moves job costs by a few per cent and the effort is better spent elsewhere.
The analysis is worth doing once to find out which situation you are in.
The operation that has changed underneath its base
The usual history: the base was chosen when the work was manual, the work has since been part-automated and part-outsourced, and nobody revisited it.
A base is a model of what drives cost. Models go out of date in the same way that shrinkage figures and driver coefficients do, and for the same reason: nothing in the routine prompts anybody to check.
Setting a review interval
Three years, or after any material change in how the work is done. More often is unnecessary — bases should be stable — and never is what currently happens.
The review is the test described above: recompute, compare rankings, decide. A morning every three years against a model that silently decides which work looks worth having.
The smallest useful correction
Adding a flat per-job element alongside the hours base, covering the transaction-driven part of the pool. One extra number, computed from job count.
It is the change that most often reorders the margin ranking, because it is the only one that distinguishes a two-hour job from a hundred- hour one in anything other than proportion.
The sentence to publish
"Overhead is recovered at £X per direct labour hour, from a pool of £Y, on a base of Z hours. The base is labour hours because the pool is predominantly people-driven; this was last tested in March."
Four facts and a justification. The justification is the part that makes it reviewable rather than inherited.