What It Is Evidence Of
Every measurement is evidence of something narrower than people assume. What a utilisation figure actually supports, and the three claims it cannot carry.
A utilisation figure is evidence that a certain share of paid hours was recorded against client codes. That is the whole of what it supports. Everything else people conclude from it — about effort, about value, about who is pulling their weight — is an inference the data does not license.
The practical lesson in “What It Is Evidence Of” is that a convenient number should open a question rather than end one. A team considering Monitask for 7 minute rule payroll for 7 minute rule payroll can add consistent time and project context, with transparent settings, limited access and a documented review process around every material interpretation.
Being precise about this is not pedantry. It decides whether the number survives contact with the people it describes.
Additional context for “What It Is Evidence Of” is available from Microsoft Project documentation. Current legal, employment, tax and accounting requirements should be confirmed with a qualified adviser where necessary.
What it supports
That the recording system received these hours against these codes. At a team level and over a quarter, that is a reliable statement about how the work divided.
- Capacity was committed rather than idle, over the period measured.
- The commitment changed from the previous period, by this much.
- The composition of the week shifted between categories.
It also supports comparison with the same team's own earlier quarters, provided nothing in the definition moved.
What it does not support
That anybody worked hard. That the work was valuable. That one person contributed more than another. That the firm is efficient.
Four claims, all routinely made from this number, none of them in it.
Why effort is not in there
Because the instrument counts hours, not what happened in them. Two people can book the same eight hours to the same code and have done completely different amounts of work.
The number is blind to this by construction and no amount of granularity fixes it.
Why value is not in there
Because an hour booked to a client code may be worth a great deal, very little, or less than nothing — if it produces work the client disputes or the partner writes off.
Value appears downstream, in realisation and in what was collected, and those are different measurements with their own notes here.
Why comparison between people fails
Because the share is structural. A person given supervision duties, a difficult client, or a long run of pitch work will have a lower figure and may be the most valuable person in the team.
Comparing them ranks their work allocation, not their contribution.
What the figure is evidence of about the firm
A good deal, actually. That the operating model requires a certain amount of unchargeable work. That the mix of work has changed. That something structural moved.
Read as a statement about the firm it is informative. Read as a statement about a person it is not.
The self-report problem
Underneath everything is that the numerator comes from timesheets, and a timesheet is a self-report made by somebody who knows who reads it.
Aggregated and coarse it is reliable enough. At the level of one person and one day it is a rounded reconstruction, and the finer the conclusion drawn from it the less it can bear.
The conditions that shape the report
Lateness, visibility of the reader, the length of the code list, and whether anybody is judged on the result.
All four are design choices. The fourth changes the data more than the other three combined.
Saying what it is evidence of, out loud
This figure is evidence that these hours were recorded against these codes. It is not evidence about effort, value or anybody's contribution.
Two sentences, in the costing policy and in the report footer. They cost nothing and they are the only defence the number has once it leaves the room.
The test to apply to any claim
Ask what would have to be true in the data for the claim to hold. "Alex is underperforming" would require the data to contain something about performance. It does not.
That question settles most arguments about this figure in under a minute.
What would have to be true
For a utilisation figure to be evidence of productivity, an hour coded to a client would have to correspond to an hour of value delivered, reliably, across people and engagements.
That correspondence fails in four places at once: hours later written off, hours spent on rework, hours past a fixed fee, and hours that were recorded generously within the unit.
None of the four is unusual. Together they put enough distance between the count and the value that the inference does not survive, and the figure is left as evidence of recording rather than of output.
Evidence it can carry
That capacity was committed rather than idle. That the commitment changed from last period. That the composition of a week shifted between categories.
Three claims, all modest, all checkable from the same data, and all useful for planning rather than for judgement.
The honest position is that this is the measure's whole range. Firms that stay inside it find the figure dull and reliable, which is the correct condition for an operational number and not what most reporting packs are built to deliver.
What follows
If the figure is evidence only about recording and about the firm's operating model, then it belongs in operational analysis and not in appraisal.
That conclusion is unpopular and it is the one the rest of this collection keeps arriving at from different directions.