Timesheet completion

What usually decides this number is WHEN logging happens rather than whether people are willing to do it. Time written down inside the week it was worked is an act of recording. Time written down at the end of a period is an act of remembering, and the same work produces a lower and shakier number when it goes through memory. That is the mechanism this metric most often reflects, and it is stated here as the likely reading rather than as a finding: confirming it for your team means comparing when hours were worked against when they were entered, entry by entry, and this page does not read that. What can be said without it is narrower and still useful. A completion share short of its goal is rarely a sign that the work did not happen, and nothing about a tracker being broken shows up in this number at all.

Nudge this week rather than reconstructing at period close. Ask for the missing days while the week is still recent enough to be recorded rather than recalled.

A gap in this share is usually concentrated in a few people running days behind, not spread evenly across the team, so a firm-wide policy response costs everybody time to fix something most of them are not doing. It is also the cheapest possible correction while the week is recent: the same hours entered a month later are an estimate, and every number computed from them inherits that. This page cannot yet name who is short, which is what the section above says would take a per-person read, so the ask is for the people your own tracker already shows as behind.

This closes no measured driver above, and does not claim to.

Move the close inside the week. A short, standing end-of-week close with the tracker open beats any amount of month-end reconstruction, because it asks people to record rather than to recall.

Two structural pieces do most of the work here. First, logging becomes part of finishing a week rather than a separate administrative task, which is the difference between a habit and a chore. Second, onboarding names time logging as part of the job from the first day, so a new starter never learns the slower pattern in the first place. Both are changes to the ritual, and the product cannot make either of them for you.

This closes no measured driver above, and does not claim to.

This is the metric other metrics are made of, so ignoring it does not leave one number wrong. Utilization, billable share and every per-client margin on this board are computed from logged hours, and each of them inherits the missing time as a quiet understatement. The damage compounds in a specific way: unlogged hours look like capacity, capacity looks like room for more work, and the firm sells against room it does not have. By the time it is reconstructed at period end, the reconstruction is a guess wearing a ledger's clothes, and it is a guess that has already been billed against. Where this lands is not this tile. It is Utilization reads 50%, with no band set for it. There is no band set on that tile, so it can show you the movement and cannot tell you whether the movement is already too far. Utilization is computed from the same logged hours, so it inherits every hour that never got written down. That is why the damage is visible there before anybody notices it here.