Capacity utilisation
Capacity utilisation measures logged hours against the hours a person was really free to work, once leave, holidays and other known absence come off. Somebody who took two days off in a five day week is judged against three, so a light week does not read as a weak one.
The narrower baseline fixes an unfairness. Judge a week against contracted hours and anyone on holiday looks idle. A team's figure then drops every August, for reasons that have nothing to do with the team.
Take known absence out first and the shares compare across people and months. That is what makes the measure fit for planning, rather than only for reporting.
A second thing surfaces too. Once absence is gone from the baseline, a low figure means spare time nothing was given to. That is a staffing question, not an effort one.
The subtraction has to be automatic. Keep it by hand and leave approved late never reaches the sum. The figures drift from what happened. Nobody spots it, because a share carries no sign that the number under it is stale.
Say which version you are quoting. Two people can report the same team's month and differ by ten points, purely on what each took out first. Neither is lying. They are answering two different questions with one word.
Where this shows up in PeopleMuster
PeopleMuster's capacity grid already subtracts leave and holidays, marking those days L and H, so the planned hours a person is measured against exclude days nobody was ever available.
Questions people ask
What gets subtracted from capacity?
Approved leave and company holidays at minimum. Some teams also remove known non-project commitments. Whatever the list, apply it consistently or the percentages stop comparing.
Why does capacity utilisation look higher than plain utilisation?
A smaller denominator. Removing absence from the available hours raises the percentage without anybody logging a single extra hour.
Can capacity utilisation exceed 100%?
Yes, when somebody logs more than they were available for. That is worth reading as a warning rather than an achievement, especially if it repeats.