Revenue per Chargeable Hour
Yield is revenue divided by chargeable hours. It moves in the opposite direction to utilisation often enough to be worth watching.
What an hour of client work was worth to one team last quarter
£96.40
Numerator: Fee income collected in the quarter
Denominator: Chargeable hours recorded in the same quarter
Collected rather than invoiced, so the figure lags with collection.
The same team's figure two years earlier was £128.70, on 11% fewer chargeable hours. Utilisation rose across those two years while the worth of an hour fell by a quarter, and both reports were correct. One firm's own quarter, not a benchmark.
Yield is collected fee income divided by chargeable hours: what an hour of client work was actually worth.
The commercial question in “Revenue per Chargeable Hour” needs a clean distinction between time recorded, time approved, time billed and money received. Teams evaluating see the product setup in context for employee time tracking can strengthen the work-record part of that chain while leaving invoice totals, tax and settlement to the appropriate finance system.
It is a single number, it is easy to calculate from data every firm already has, and it frequently moves the opposite way to utilisation.
As a separate benchmark for “Revenue per Chargeable Hour”, see Xero accounting resources. External guidance is most useful when it prompts a specific check rather than replacing case-by-case judgement.
Why it moves the other way
Because most of the ways to raise utilisation add hours rather than revenue. Overrun, rework, deferral of improvement work, longer delivery from stale templates.
Each raises the denominator of yield while leaving its numerator alone.
The arithmetic of the divergence
A team that bills the same work in 1,200 hours rather than 1,000 has raised utilisation by twenty per cent and lowered yield by seventeen.
Both figures are correct and only one of them is on the dashboard.
What a falling yield usually means
Work taking longer. Discounting. A shift toward lower-rate work. Or the compounding deferral described in the behaviour section.
Four causes, distinguishable by looking at rate mix and hours per engagement separately.
Why it resists the individual responses
Yield cannot be improved by recoding an hour. Recoding an ambiguous hour to a client code raises the denominator and lowers yield.
So the eight responses in the behaviour catalogue either do nothing to it or make it worse, which is exactly what a usable measure should do.
What it cannot see
Whether capacity is being wasted. A firm doing very little work very profitably has excellent yield and a serious problem.
Yield needs utilisation beside it for the same reason utilisation needs yield.
The three-number report
Chargeable hours, available hours, collected revenue. Everything else in this collection is derived from those three.
Utilisation is the first over the second, yield is the third over the first, and the product of the two is revenue over available hours.
Revenue per available hour
The single figure that combines everything: collected fee income divided by the firm's available hours.
It cannot be improved by recoding, by deferral, by overrun or by rounding. It can be improved by selling more, delivering faster, or pricing better.
Why it is rarely used
Because it does not decompose into a person. It is a firm or team figure, and firms want an individual number.
That is a reason it is better rather than worse, and it is the measure the last section of this collection ends on.
Calculating yield by engagement
More useful than the average. Collected fee divided by hours, per engagement, sorted.
The bottom of that list is usually a short set of jobs that consumed a large share of hours, and it answers questions no ratio can.
The pattern in the bottom of the list
Fixed-fee work that overran, long-standing clients on old rates, and work taken to fill the bench.
All three are visible in the yield table and invisible in the utilisation report, and all three are commercial decisions somebody can revisit.
The seniority adjustment
Yield varies with grade mix, because senior hours bill higher. A team that shifted work down the pyramid will show falling yield and rising margin.
So yield needs the mix beside it, as everything in this collection needs its context beside it.
The two sorts of falling yield
A fall caused by the rate: discounting, old rate cards, a shift toward cheaper work. Visible in the rate mix.
A fall caused by the hours: overrun, rework, slower delivery, stale internal assets. Visible in hours per engagement.
The two have nothing in common except their effect on the figure, and distinguishing them takes one additional cut of the same data. Firms that do not make the cut address whichever cause they happen to assume.
What rising yield can conceal
A firm that has stopped taking difficult or developmental work will show improving yield and a narrowing pipeline.
So yield is gameable at the portfolio level even though it resists gaming at the timesheet level: the route is not recoding an hour but declining a category of work.
That is a slower and more visible manoeuvre, and it is caught by looking at volume beside yield, which is the reason the four figures are read together rather than ranked. Every measure in this collection has a blind spot, and the practical arrangement is not to find one without a blind spot but to put enough of them on the page that the blind spots do not overlap.
Where to put it on the page
Beside utilisation, not under it. The two are read together or the divergence is missed.
A page that shows utilisation prominently and yield in an appendix has the same blind spot as a page that omits yield entirely, because appendices are not read and divergence is only visible when both figures are in view.
Adjacency is the requirement, and it is a layout decision.
What to do first
Calculate yield for the last eight quarters. It takes an hour with data the finance system already holds.
Where it has fallen while utilisation rose, the firm has been reporting activity and losing value, and the divergence will usually be larger than anybody expected.