What Time Data Cannot Tell You

A system that answers everything asked of it is answering some things badly. Defending the boundary is what keeps it trusted on the rest.

Whether someone is working hard

Why not: hours measure duration, not contribution, and the data is not accurate enough to compare individuals in any case.

Use instead: the work itself. Output, quality, and the judgement of someone who sees it.

Whether someone was actually working

Why not: a record of duration says nothing about what happened inside it, and the systems that do claim to answer this measure keystrokes rather than work.

Use instead: the same as above. If this question is being asked seriously, it is a management problem that surveillance will not solve.

Why a project cost what it cost

Why not: the timesheet records where effort went, not why it went there.

Use instead: the effort data as the starting point, then ask the people involved. The reason is almost always in the scoping or the brief.

Whether an estimate was reasonable

Why not: the actual tells you what happened, not whether it was avoidable.

Use instead: the estimate-to-actual factor across many instances, which identifies a systematic pattern rather than judging one case.

Who is most productive

Why not: the measure would be hours or billable percentage, both of which reflect the work assigned, and both of which are gameable.

Use instead: nothing from this system. It is the question this whole collection argues against answering from time data.

What the true cost of the organisation is

Why not: unbilled and internal work is systematically under-recorded, so total effort is understated by an unknown margin.

Use instead: the recorded figure with the known omission stated, plus one hand-reconstructed project as a calibration point.

Whether the record is accurate

Why not: the system reports what was entered.

Use instead: the four data quality checks — delay, tidiness, catch-all rate, completeness — which are proxies and are the best available.

How to decline

Say what the data does support, which is usually adjacent and substantial.

Say what would answer the question and roughly what it would cost.

Write the limitation into the report, not the covering email, because the report circulates and the caveat does not.

Do not produce a precise figure you cannot support. Reporting effort to two decimal places on data reconstructed weekly is a statement about the report rather than the work.

What it does support

Where effort went, by project and by activity type.

How long work takes, against what was estimated.

What was billed against what was recorded.

Whether people are working more hours than they should be.

Four things answered usefully is a system worth running, and defending the boundary is what keeps those four trusted.

Precision that the method supports

A discipline that keeps the reporting credible.

Weekly reconstruction supports statements about shape and trend.

It does not support two decimal places, or comparisons between individuals, or confident differences of a few percent.

Round the reporting to the precision the data carries, which is usually coarser than the system displays.

Say what the band is.

A figure reported more precisely than it deserves is the first thing an informed reader tests, and the whole report goes with it.

Writing the caveat where it survives

A limitation stated in conversation is dropped by the second retelling.

Put it in the report, in the section people read.

Attach it to the number itself: the method, the delay figure, the known omission.

Repeat it in any summary, because summaries circulate and originals do not.

State what was not measured.

In this field the consequence of a confident wrong figure lands on a person's pay or reputation, which is why the caveat is not a formality.

A concrete product reference

When translating this principle into a buying test, read the feature details provides a concrete feature and workflow reference. Verify the relevant behaviour in a trial, retain the exported evidence and judge it against the purpose and limits described above.