Getting timesheets done without nagging

Nobody dislikes timesheets in principle. They dislike reconstructing a fortnight from memory on a Friday afternoon, for a number they believe is only used to bill someone.

Fix the friction before the process

• Make the list of things to log short and relevant: this person's projects and tasks, not the whole organisation's.
• Allow logging as work happens, not only at week end.
• Pre-fill from evidence where it exists — timers, meetings, assigned work — and let people correct it.
• Keep the weekly submission to a few minutes; anything longer gets deferred, and deferral is the whole problem.

Weekly, not monthly

Monthly capture means the cost position is up to four weeks stale and the estimates are unimprovable. Weekly capture is close enough to reality that delivery can use it while the project is still running.

Explain the second purpose

Logged effort is what makes cost-to-date, forecast cost and margin real. It is also the only honest feedback on estimates: without it, next quarter's proposals are priced on the same optimism as last quarter's. Teams treat time capture differently once it stops looking purely administrative.

Approve for signal, not ceremony

Approval should catch the entries that look wrong — effort against the wrong project, a week far above or below the norm — rather than audit every line. Locking approved time matters because the commercial numbers read from it; the review itself should stay fast.

One measure of whether it is working

How current is the cost position on a Wednesday? If the answer is "last month", the problem is the capture rhythm, not the people.