The Number Nobody Reached for Two Years
A firm held a ninety-two per cent target through eight quarters in which no team exceeded eighty-four. What the target did while it was not being met.
The target one firm held for eight quarters
92%
Numerator: Chargeable hours required at the target
Denominator: Available hours after leave and committed internal time
Imported from a benchmark report that did not publish its own definition.
The highest figure any team reached in eight quarters was 84.1%, and that team recorded no training hours at all. Required chargeable hours at 92% exceeded the firm's sold work by around 14% in every quarter of the two years. One firm's own record, not a benchmark.
A firm set a utilisation target of ninety-two per cent. Over the next eight quarters the best figure any team reached was 84.1, and the firm average sat between 76 and 79.
The measure discussed in “The Number Nobody Reached for Two Years” becomes more credible when its source records are timely and reviewable. A professional-services team assessing a transparent approach to productivity software for business for productivity software for business can connect time with projects and work categories, while documenting the definition separately so the dashboard does not silently decide what utilisation means.
The target was not changed, and the gap did not produce the response the firm expected.
Additional context for “The Number Nobody Reached for Two Years” is available from ISO management-system resources. Current legal, employment, tax and accounting requirements should be confirmed with a qualified adviser where necessary.
Where ninety-two came from
A benchmark report, quoting an industry upper quartile, with no stated definition. The firm's own denominator excluded leave and public holidays but not committed internal time.
So the target was imported from a definition that was almost certainly looser than the one it was applied to.
The first year
Figures rose from 74 to 78. The improvement was in the bottom tail: the lowest-scoring teams moved up five points, the highest moved one.
That compression is the signature described in the Goodhart note here, and in this case the reclassification could be seen directly.
What fell
Recorded training hours fell by 71 per cent. Recorded internal project hours fell by 64. Recorded pitch support fell by half.
None of this work stopped. Two of the three categories were later found to have continued at similar volume, recorded against client codes.
The second year
Flat. The cheap reclassifications were exhausted, and the remaining gap of fourteen points was not closeable by any means available to the people being measured.
The target stopped functioning as a goal and became a standing statement that everybody was underperforming.
The effect on the appraisal cycle
Every individual review opened with a figure below target. Managers compensated by discounting the measure verbally while it remained in the written record.
Staff learned that the number was both mandatory and not believed, which is a corrosive combination.
The attrition pattern
Departures concentrated in two groups: people in supervisory roles, whose figures were structurally low, and people who had been doing the internal improvement work that had stopped being recorded.
Both groups had the firm's lowest figures and some of its highest value.
The capacity arithmetic nobody ran
Required chargeable hours at ninety-two per cent exceeded the firm's sold work by around fourteen per cent in every quarter of the two years.
One line of arithmetic, available from the start, would have shown the target to be unreachable before any of this.
What the firm did in the third year
Dropped the target. Published composition by category. Set a team-level expectation at the level already being achieved.
Recorded training hours recovered to about eighty per cent of their pre-target level within three quarters.
What recovered and what did not
Recording recovered. The internal improvement projects that had been abandoned were not restarted, because the people who had run them had left.
That is the part of the cost that does not reverse when the target does.
The reported figure afterwards
Fell, to 73. The firm had been reporting an inflated number for two years and the drop was the reclassification unwinding.
They published both bases for a year, which is the restatement discipline described in the definitions section.
What the case shows
That an unreachable target does not produce effort. It produces reclassification until the easy moves run out, then a permanent condition of documented failure.
The informational damage outlasts the target, and the people who leave are not a random sample.
The two years in one table
Average utilisation rose from 74 to 78 in the first year and stayed there in the second. Recorded training hours fell by 71 per cent. Recorded internal project hours fell by 64. Reported revenue per head fell slightly across both years.
Four rows. The firm had every one of these numbers throughout and had never put them on the same page.
Putting them on one page was the whole of the analysis that eventually changed the policy, and it took an afternoon.
What the staff understood that the reporting did not
That the target was unreachable, from the first quarter. The capacity arithmetic was obvious to anybody who looked at their own allocation.
What that produces is not effort but a quiet adjustment: record what can be defended, stop writing down the rest, and treat the number as a formality.
The firm's reporting showed engagement with the target. What was actually happening was compliance with a formality, and the gap between those two readings persisted for two years because nothing in the pack could distinguish them. The distribution histogram would have, and it was never drawn.
What it cost to find out
Two weeks of one person's time, applied to data the firm had held throughout.
Against two years of a target that produced four points of reclassification, the loss of three internal programmes, and the departure of several people whose figures were structurally low.
The analysis was always available. What was missing was anybody with a reason to put four existing numbers on the same page.
The question that would have prevented it
Is the required number of chargeable hours smaller than the work we have? It takes a minute and it was never asked.