A healthy utilisation rate for a UK agency delivery team is 70 to 80% of available hours, and I plan capacity at 75%. Agency-wide, once you include the owner, ops and anyone selling, the UK average sits between 60 and 65%. Below 60% for a delivery team means you are paying for hours you cannot sell. Above 85% means quality, training and new business are quietly funding your billable number. You will see it spelled utilization rate in the US tools; the maths is identical.
Those are the benchmarks. Now the story behind why I care about this number so much.
An owner I coach told me his team was maxed out. Completely full, no room for another client, and hiring felt like the only way forward. So we went looking for the numbers behind that, and there were none to find. No logged time, no view of where the hours went, just a feeling that everybody was busy and everybody was great. He was certain, and to be fair, so was I when I ran my own agency in Belfast on the same instinct.
Utilisation is the number that replaces that feeling. Get it measured and it changes three decisions at once: when you sell, when you hire, and what you charge.
What is utilisation rate for an agency?
Utilisation is billable hours as a percentage of available hours. Two definitions inside that sentence do all the damage, so pin them down first.
Billable hours are hours logged against client work that you charge for. Available hours are the hours you actually pay for after holidays come out, and this is where the calculation usually goes wrong. A person on a 37.5-hour week does not give you 52 weeks of capacity. Take off statutory holiday and bank holidays and you are left with roughly 46 working weeks, which is about 1,725 available hours a year, or 144 a month.
Run that through a real team. Five people doing client work gives you a monthly pot of around 719 hours. If the logged billable time last month was 470 hours, your delivery utilisation was 65%. That is the whole calculation, and the calculator further down does it for you.
One distinction worth keeping: measure your delivery team and the whole agency separately. Your designers and developers should sit far higher than the blended number, because the blended number carries you, your ops person and everyone whose job is keeping the machine running rather than billing. Mixing the two produces a figure that flatters nobody and informs nothing.
What is a good utilisation rate for a UK agency?
The UK benchmarks are fairly consistent across the agency ops platforms and the accountants who publish agency data. Here is how I read the bands for a delivery team:
- 75 to 85%: strong. Designers, developers and copywriters live here when scoping is tight and admin is kept off their plate.
- 70 to 75%: healthy. This is the planning zone, and 75% is the number I build capacity models on.
- 60 to 70%: below the band. Common, survivable, and expensive. Ten points of utilisation on a five-person team at £90 an hour is roughly £77,000 a year.
- Under 60%: a problem. Either there is not enough sold work, or the work is there and the hours are leaking into unbillable scope.
Account managers and strategists sit lower, around 65 to 75%, because relationship time and internal coordination are real parts of the job. And the blended agency-wide average across the UK lands at 60 to 65%, which surprises owners who assume everyone else is running hotter.
The top end needs a warning of its own. A delivery team above 85% for a sustained stretch is not a triumph. There is no slack for a pitch, no time for training, and the quality slips in ways clients notice before you do. Fine for a month. Corrosive over a quarter.
Work out your own number
Agency utilisation calculator
Logged billable hours against your team's true available hours, with the gap to a 75% plan priced at your rate.
Below the healthy band. The gap is worth pricing before you decide you need another hire.
Two honest notes on what the calculator tells you. The gap it prices is fee capacity, hours you are already paying for that could carry client work at your rate. Treat that figure as the size of the prize rather than booked revenue. And the output is only as good as the billable-hours figure you type in, which brings me to the uncomfortable part.
Why 100% is the wrong target
Every founder who has done the maths on their payroll has felt the pull of this: if we pay for 719 hours, we should bill 719 hours. I understand the instinct and I am telling you to drop it. Around a quarter of any honest working week goes on things that keep the agency alive without ever touching an invoice. Internal reviews, proposals, training, the thinking time that makes the billable hours worth paying for.
That is why I plan at 75% effective utilisation, and why I give every owner I coach the same number. Building the slack into the plan means the plan survives contact with a real week. Chasing 100% just guarantees your plan is fiction, and fictional capacity is how growth crashes into delivery.
The point of the number was never to squeeze people. Utilisation exists to make selling a decision backed by data. When you know the team has 140 spare hours next month, you can go to market for exactly that: one more retained client and a small project. Full teams and empty teams both come from selling blind.
It sets your hiring trigger too. Two directors I coached were stuck in the part-time-or-full-time debate, staffing gaps on one side, a maternity leave on the other, no way to settle it. We threw out the question and replaced it with a threshold: when the delivery team holds around 80% of capacity for three months running, hiring starts. The panic left the decision, because the trigger now lives in the data rather than in whoever argued loudest that week.
Utilisation is what makes your rates real
A £900 day rate on a half-empty calendar is a £450 day rate wearing a nicer number. I made that point in the guide to what UK agencies charge, and utilisation is the mechanism behind it. Your rate card and your utilisation multiply together to produce your actual revenue per head, which is why they belong in the same conversation and on the same KPI dashboard.
Run the bridge in both directions. Backwards from the fee: a £36,000 retainer at £90 an hour buys the client 400 hours a year, or 33 a month. The moment logged time passes that, you are donating margin, and you will only ever catch it if the hours are tracked per client. Forwards from capacity: if the team is at 75% and you are winning around 70% of the proposals you send, the constraint has stopped being sales. You are too cheap. Raising rates at full utilisation is how agencies grow revenue without adding a single head.
How to get a number you can trust
None of this works from memory or gut. It works from logged time, and there is a way to do that without it feeling like surveillance, which is the fear that stops half the owners I raise it with.
Put the timer inside the task itself, a start and stop on the job in whatever project tool you run, so logging is frictionless and skipping it is harder than doing it. Frame it to the team as exactly what it is: data about the business, never a performance score. Then let three truths surface. Real utilisation per person. Time-per-task against what the task should take. True profitability per client, which is where the over-serviced accounts finally show themselves.
Expect surprises. Logged time has a habit of turning up things like an hour a day quietly vanishing from someone’s calendar, and scoped-at-ten-hour tasks that have been taking sixteen since January. When a gap appears between what a task takes and what it should take, ask one question before acting: is this a skill issue or a will issue? The answers lead to different conversations.
One more reason to build the habit now: every efficiency gain you make only becomes real once you can measure it. If AI tooling cuts a production task by 10%, that is 10% of that task’s hours back in the capacity pot, ready to be sold. Without time data it is just a nice feeling.
Where to start
This week, calculate the crude version. Contracted hours, minus holidays, against whatever billable record exists, even if the record is rough. The calculator above takes five minutes and gives you a baseline to distrust properly.
Over the next 30 days, get per-task, per-client time logging live for the delivery team, framed as data-gathering from day one. Do not act on week one’s numbers; early logging always wobbles.
Next quarter, set the two thresholds and write them down. Plan capacity at 75%. Start hiring conversations when the team holds 80% for three months. Then point your sales effort at the spare hours the data says you actually have.
If you want to see where utilisation sits among the six pillars that decide whether your agency can grow without breaking delivery, the Agency Health Scorecard takes two minutes. If the numbers in this guide made you uncomfortable in a useful way, that is the next step I would take.