A good agency utilisation rate is set by role

Start from the hours a person really has

Utilisation is logged hours over available hours. Get the bottom half wrong and every target you set on top of it is wrong too.

A full-time year looks like 52 weeks of 40 hours, which is 2,080 hours. Nobody works all of them. Take out your company holidays and each person's leave, and a lot of that year disappears before a single client asks for anything.

Try it with your own calendar. Ten holiday days and fifteen days of leave remove 200 hours at eight hours a day, so 1,880 hours are left. That's the number your target is a share of, and it's different for someone who joined in March.

Why one number for the whole agency misleads you

An agency-wide rate averages people whose weeks look nothing alike. A developer on one client build can log most of the week to that project. A tech lead reviews code, sits in client calls and helps hire, and a lot of that time sits outside any single project.

Put both into one average and a healthy number can hide a person running at well over their expected day. You'd see the average and miss the burnout.

So set the target per role and read it per person. Your leads get a lower project share than your developers, because their other work is still work. A rate you can't hit without skipping reviews or one-to-ones isn't a target. It's overtime you've planned in advance.

What the published figure tells you

Published utilisation benchmarks are scarce, and most of the figures you'll find quoted online have no study behind them. Deltek's Clarity A&E study, released on 12 May 2026, found staff utilisation just under 60% across the architecture and engineering firms it surveyed.

Those firms bill time the way many software agencies do, so the figure is a fair sense check. It isn't a target for your team. Your mix of roles, your client contracts and your leave pattern set that, and you already hold all three.

Measure against the expected day, not a fixed 40

PeopleMuster calculates utilisation from the hours your people log each day against each project in the daily recap. Each person's week shows total logged hours against expected hours, with a completion percentage.

Your utilisation report gives you total hours, a utilisation percentage and average daily hours. It breaks them down by project, by department and by sheet status, with a week-over-week view and a weekly trend. Send it to Slack on the schedule you choose, and your leads read it where they already talk.

Your planning grid uses the same expected day. Bands for under, partial, full and over follow the working day you set, so a team on a shorter standard day isn't marked as under-used. Approved leave shows as L on those days and company holidays show as H, which is how the available hours in your rate stay honest.

Signs your target needs changing

A target is a planning decision, and it should move when your team does. Check it against these at the end of each month.

• A few of your people sit above their expected day week after week.
• Your leads hit their target only in weeks with no hiring or reviews.
• Logged hours jump just before a deadline and fall right after.
• A department's rate looks fine while one project inside it runs hot.
• Nobody can say which hours in the rate were approved and which were still drafts.