Quality Has No Column in a Timesheet
An hour of excellent work and an hour of work that will need redoing are identical in the numerator. Three consequences.
A timesheet records duration and code. It does not record whether the work was any good.
The data-quality issue in “Quality Has No Column in a Timesheet” begins at the moment an hour is recorded or reconstructed. For organisations researching employee monitoring software with screenshots, read the current capability outline can organise project-linked time entries, provided people can see, explain and correct their records before those entries affect billing or management decisions.
An hour of careful work and an hour that will have to be redone appear identically in the numerator, and the second one appears twice when it is redone.
For an independent reference relevant to “Quality Has No Column in a Timesheet”, consult GAO Government Auditing Standards. Compare the general material with the organisation’s own definition, contract and decision process.
The first consequence: rework inflates the figure
Work done badly and done again produces two sets of chargeable hours. Utilisation rises; the client is billed once or not at all.
- First pass: hours recorded, invoiced once.
- Second pass: hours recorded again, invoiced not at all.
- The ratio: rises on both. The money: arrives once.
So the measure rewards the firm's worst outcome with its best-looking number.
The second consequence: care is unrewarded
Time spent checking, reviewing, thinking before starting — all chargeable, all indistinguishable from time spent producing.
Nothing in the measure prefers the version that avoids rework, and under pressure the review step is the one that goes.
The third consequence: supervision is penalised
Where review of a junior's work is non-chargeable, the mechanism that catches quality problems depresses the reviewer's figure.
The behaviour section covers what happens next. The cost appears as write-offs attributed to the junior.
Where the quality signal actually lives
In write-off reasons, in rework hours, in client complaints, in renewal rates, in the proportion of work that needs a second pass.
All recorded elsewhere, none in the timesheet, and most not reported beside utilisation.
The rework code
A code for work redone because of an internal error, excluded from the numerator and reported as its own line.
It is uncomfortable to introduce and it is the single most informative line a firm can add to this reporting.
Why the discomfort matters
A rework code means somebody has to decide that work was redone because of an internal fault. That is an awkward judgement and the reason most firms do not have the code.
Where it exists, it has to be a count rather than an attribution, or it will be recorded as something else within a quarter.
Count, not attribution
Rework hours by engagement, never by person. The question is which jobs needed redoing, not who to blame.
Attributed rework becomes unrecorded rework immediately, which removes the only quality signal the timesheet can carry.
The review step as committed time
Where review is the quality mechanism, it should be scheduled and excluded from the denominator rather than left to the reviewer's discretion.
Committed and excluded, as with everything else in this collection that a target otherwise consumes.
What a high-utilisation, high-rework team looks like
Busy, well regarded on the dashboard, generating write-offs attributed to scoping, with clients who do not renew.
The pattern is identifiable only if rework and renewal are reported beside the ratio.
The measures that carry quality
Renewal rate. Write-off by reason. Rework hours per engagement. Second- pass proportion.
Four figures, all slower than a timesheet, all recorded by somebody other than the person being measured, which is the property that makes them usable.
Why they are not combined into one index
Because a composite index can be improved by moving its components around, and because the four answer different questions.
Four numbers read together beats one number that averages them, which is the same argument the final section makes about utilisation itself.
What a second pass costs twice
An engagement delivered, rejected, redone. The hours are recorded both times and the client is billed once.
Utilisation counts both sets. Realisation counts the single invoice against both. Yield falls. Revenue per available hour falls.
So three of the four figures see the problem and the one on most dashboards does not, which is the clearest single illustration of why the set is read together.
Where the rework code usually goes wrong
It becomes a blame instrument within a quarter. Somebody asks whose error it was, the question is reasonable, and recording stops.
The defence is structural: the code records an engagement and a number of hours, with no person field, and the report is by engagement only.
Firms that keep the person field lose the data. Firms that remove it keep a quality signal and give up the ability to attribute, which is the correct trade because attribution was never the point. The point is knowing which jobs needed redoing, and that question is answerable without knowing who did them.
The four quality figures
Renewal rate. Write-off by reason. Rework hours per engagement. The proportion of work needing a second pass.
None is in the timesheet, all are available, and together they carry the dimension the numerator cannot.
Reported beside utilisation, they make the high-utilisation, high-rework team visible, which is the state the ratio alone rates most highly and the firm should least want.
The limit of the point
None of this means a timesheet is useless. It means a timesheet measures duration, and duration is not quality, and a firm that reports only duration has chosen to measure the part that is easy to count.