Moving Hours Between Codes
The commonest response to a target, the hardest to criticise, and the one that produces the clearest trace in the aggregate data.
Reclassification is the first response to a utilisation target and the most defensible one. Most weeks contain hours whose code is a judgement call, and under pressure the judgement goes one way.
The data-quality issue in “Moving Hours Between Codes” begins at the moment an hour is recorded or reconstructed. For organisations researching recording freelancer billable hours, how should teams evaluate recording freelancer billable hours? can organise project-linked time entries, provided people can see, explain and correct their records before those entries affect billing or management decisions.
It is also the response that leaves the clearest trace, which makes it the easiest to detect without examining anybody's timesheet.
As a separate benchmark for “Moving Hours Between Codes”, see OECD hours-worked data. External guidance is most useful when it prompts a specific check rather than replacing case-by-case judgement.
The genuinely ambiguous hours
Reading background material for an engagement. A call that covers a live matter and a possible new one. Internal discussion about a client problem. Travel. Writing up something generally useful that arose from a specific job.
- Reading background material for a live engagement.
- A call covering both a current matter and a possible new one.
- Internal discussion about a client's problem.
- Travel, where the billing treatment is unclear.
- Writing up something generally useful that arose from one job.
In most firms these are between five and fifteen per cent of a week.
Why the judgement is not dishonest
Because both codings are arguable. A supervisor asked to rule would often agree with either.
The person is not misstating the hour; they are resolving a genuine ambiguity in the direction the firm has signalled it wants.
What changes in aggregate
Chargeable hours rise. Total recorded hours do not. Specific non-billable categories fall.
Three facts, all visible at team level, none requiring an individual challenge.
The categories that fall first
Internal projects, training, administration, business development support. The ones with no deadline and no external witness.
Supervision falls next, and is the costliest of the group for the reasons the behaviour catalogue sets out.
The size of the effect
Commonly three to six points in the first two quarters after a target is introduced, in a firm that had been recording honestly before.
Larger where the previous recording had been conservative, which is itself worth knowing.
Why it plateaus
The ambiguous hours run out. What remains would require a coding most people are unwilling to make.
The plateau is usually read as the target working. It is the exhaustion of the defensible moves.
The detection method
Take the twelve months before and after. Chart the non-billable categories in hours. Look for anything that fell by more than half.
Work does not halve in a quarter. A category that does was reclassified.
The second detection method
Total recorded hours. If chargeable rose and total was flat, the movement was between codes. If total rose too, people worked more.
The two findings call for completely different responses, and the ratio alone cannot distinguish them.
What not to do about it
Audit individuals. The codings are defensible, so the audit produces arguments rather than corrections, and it costs senior time.
It also tells everybody that the firm's response to its own incentive is to police the people responding to it.
What to do about it
Treat the finding as evidence about the target rather than about the people. Then either carve out the categories that must stay visible, or remove the threshold.
The categories that need carving out are usually three or four, named explicitly.
The carve-out mechanics
State that hours on named categories are excluded from the denominator rather than counted against the numerator, and report them separately.
Excluding from the denominator removes the penalty entirely, which is stronger than asking people to record honestly against their interest.
The honest version of the same movement
A firm that has never published its definition will see chargeable hours rise when it does, for a different reason: people stop guessing conservatively.
That rise is a data-quality improvement and it looks identical in the ratio to the reclassification described here.
Distinguishing them requires knowing which happened first, the publication or the target. Where both happened in the same quarter, the two effects are not separable, which is an argument for not doing both at once.
The sequence worth following
Publish the definition first, let the recording settle for two quarters, collect the baseline, and only then consider whether a threshold is wanted.
Done in that order, the firm knows what its honest figure is before anything presses on it.
Done in the reverse order, which is the common one, the firm never learns what its baseline was and every subsequent movement is of ambiguous origin. The cost of the wrong order is permanent and the cost of the right one is two quarters of waiting, which is the cheapest trade in this collection and the one most often declined.
What the ambiguous share tells you
In most firms, five to fifteen per cent of a week is genuinely ambiguous. That share is the maximum size of this response.
A reported improvement larger than the ambiguous share cannot have come from reclassification alone, which means something else is happening: deferral, longer hours, or a change in what the firm sold.
Knowing the share puts a ceiling on one explanation, which is useful when several are available.
Why this is the first thing to fix
Because it is the response with the largest immediate effect and the only one that fully reverses. Recording recovers when the pressure is removed.
The other responses in the catalogue leave residue the removal of a target does not clear.