Agency revenue per employee benchmark: the range, how to count it honestly, and what a low number is telling you

Revenue per employee is the one agency metric a buyer, a bank and a founder all read the same way, which is why it gets quoted so often and measured so badly. Divide annual revenue by headcount and you have a number in thirty seconds. Whether that number means anything depends on what went into the numerator and who got counted in the denominator.

The benchmark range

Hydra works to $150,000 to $200,000 of revenue per full-time employee for a services agency. Below $120,000 the agency is almost always carrying too many people in coordination roles relative to the people doing the delivery, and that pattern shows up before anything in the P&L does.

The external figure worth putting next to that comes from Predictable Profits' 2025 Agency Growth Benchmark, which reports on more than 300 seven- and eight-figure agencies. It puts revenue per employee at $150,000 for seven-figure agencies and $225,000 for eight-figure agencies. The published material does not describe the sample or how the figure was calculated, so treat the gap between the two tiers as the useful finding: agencies that get to eight figures do it with proportionally fewer people, usually because the average account got bigger rather than because anyone worked faster.

Other ranges circulate, from $120,000 to $200,000 in most vendor blogs to a $220,000 figure attributed secondhand to Ad Age's agency reporting. We could not trace those to a primary source and have left them out.

Count it the same way every time

Three mistakes account for most of the disagreement between agencies that compare notes.

Billings in the numerator. A paid media agency that passes £2m of ad spend through its books and keeps £300k in fees has revenue per employee of £300k on one reading and £2.3m on another. Use agency gross income, which is fees plus margin on anything resold, and exclude pass-through entirely. Every agency with media, print or influencer spend in its revenue line needs to make this correction before the benchmark applies.

Contractors missing from the denominator. An agency with eight employees and six regular freelancers is a fourteen-person delivery operation, and dividing by eight flatters the figure by 75%. Convert every regular contractor to a fraction of a full-time head based on hours billed, and count the founder at one even if they pay themselves nothing.

Year-end headcount against full-year revenue. An agency that doubled its team in the last quarter has a denominator the revenue has not caught up with. Use average headcount across the year, or run the number on a trailing three months of revenue annualized against current headcount, and label which one you used.

The revenue per employee glossary entry holds the short form of these rules if you want to hand them to your bookkeeper.

What a low number is actually telling you

Revenue per employee is a ratio, so it falls for one of two reasons, and the fix is different for each.

If the numerator is the problem, the accounts are too small. An agency with thirty clients averaging £3,000 a month has to run every account lean to hold 50% gross margin, and the coordination overhead per account does not shrink with the fee. ZeroTo1 took average contract value from $8k to $20k a month while scaling, and that lever moved revenue per employee more than any efficiency program could have, because the same account management hours now carried two and a half times the revenue.

If the denominator is the problem, the agency hired ahead of demand or hired into coordination. Check the ratio of delivery headcount to everyone else. Agencies bloat in the middle, adding project managers and account coordinators to handle complexity that a written process would remove, and each of those hires lowers revenue per employee while raising payroll above the 45% to 55% of revenue a healthy agency runs. Bench time, meaning paid hours with no client work against them, is the earliest sign, and it is visible in a utilization report months before it is visible in profit.

The check that stops it going wrong

Revenue per employee is a lagging number, and acting on it directly usually means a layoff that arrives two quarters late. Watch the two numbers that move first instead. Utilization per role tells you whether the delivery team has sold capacity or is sitting on delivery capacity nobody is paying for. Payroll as a share of revenue tells you whether the next hire is affordable at all. When utilization on delivery roles is under 60% and payroll is above 55%, revenue per employee is about to fall, and the decision is to sell into the slack or stop hiring until it is absorbed, never to hire a manager to organize the people who have nothing to do.

What it looks like when it works

Fraggell Productions grew from £23k to £110k a month over 24 months by tying every hire to pod capacity, with a defined order for which role gets added when, and held gross margin above 70% on every pod while doing it. Revenue per head rose because headcount followed revenue, in that order, instead of being added on the expectation that revenue would follow.

If you want the figure calculated the same way we calculate it for clients, with pass-through stripped and contractors counted, the diagnostic starts there and goes on to the other three constraints.

Read next

• Agency utilization rate: the target, how to calculate it, and why 100% is a warning sign
• Agency financial benchmarks: the numbers a healthy agency actually hits