Reports People Act On

Time data generates an enormous quantity of reportable material, almost none of which changes a decision.

The one page

Effort by project, this period and cumulative.

Effort by activity type, with delivery, rework and supervision visible separately.

Estimate against actual for anything that completed.

Realisation, where billing applies: recorded, billed, collected.

Unbilled effort as a proportion of the total.

Hours beyond contracted, in aggregate, with a note if it is rising.

Data quality: proportion of entries made within a day, proportion of time in a catch-all category.

Seven items. Each connects to a decision someone can make.

What to leave out

Individual hours compared between people.

Individual utilisation or billable percentages.

Timesheet compliance rates, which measure whether forms were submitted.

Anything with more precision than the data supports. Reporting effort to two decimal places on data reconstructed weekly is a statement about the report rather than the work.

Charts with twelve series, which are read by nobody.

Who gets what

Project leads: their own project, effort against estimate, where the effort is going by type.

Practice or team leads: the portfolio view, capacity, realisation by work type.

Finance: cost by project, unbilled effort, the realisation chain.

The people recording time: their own record, their team's aggregate, and what changed as a result. This last audience is almost always omitted and is the one whose cooperation determines whether any of it is accurate.

The interpretation line

One sentence per number, written by a person.

What moved, why if known, and what is being done.

Updated when it changes rather than rewritten monthly.

A number without interpretation is read as noise, and readers supply their own explanation, which is usually wrong.

Precision and honesty

State the recording method alongside the figures, because weekly reconstruction and timer capture are not the same quality of data and should not be reported identically.

Report the data quality measures on the same page, so the reader knows how much weight to put on the rest.

Where a figure is an estimate, say so.

Where a category is under-recorded — and unbilled work always is — say that too, rather than letting the number stand as complete.

Closing the loop

Report back what changed because of the data.

Including where nothing did, and why.

To the people who recorded it, first.

A time system that has never visibly informed a decision will be filled in as an administrative chore, which is a self-fulfilling arrangement and the reason most timesheet data is poor.

Stating the recording method

A line that determines how the rest of the page should be read.

Weekly reconstruction and timer capture produce different quality data.

Presenting them identically is a precision claim the method does not support.

Put the method and the delay figure on the same page as the analysis.

Where a category is known to be under-recorded — and unbilled work always is — say so rather than letting the number stand as complete.

Readers adjust accordingly, and the report survives the first challenge.

Reporting to the people who recorded it

The audience most often omitted and the one that determines accuracy.

Their own record, always available.

Their team's aggregate.

What the data showed and what changed because of it.

Including where nothing changed, and why.

First, before the management version, which is a sequencing choice that costs nothing and signals what the system is for.

A time system that has never visibly informed a decision will be filled in as an administrative chore, which is self-fulfilling.

Turn the principle into a test

For an example that can make this requirement testable, consult the integration overview. Treat the page as a starting point rather than proof: reproduce the workflow with real roles, exceptions and permissions before relying on the result.