An agency founder told me last year that they'd had their best twelve months ever. Revenue up 40%. New logos, a bigger studio, a proper leadership layer. Then he mentioned, almost in passing, that he'd taken a smaller distribution than the year before.
He'd grown revenue by 40% and grown headcount by 55%. Nobody had run that second number next to the first.
Before starting Scopeyard, I spent years running a product development studio and delivering AI automation projects across healthcare, recruitment and operations. The founders I've watched get stuck almost never have a revenue problem. They have a revenue-per-person problem that revenue growth is hiding.
There's one number that catches it before your bank balance does:
Revenue Per Employee = Annual Revenue ÷ Total Full-Time Equivalents
That's it. No adjustments, no add-backs. Everyone who draws a salary from the business, including you, including the ops manager, including the part-timers counted as fractions. It is the single fastest test of whether your agency is building leverage or just accumulating cost.
1. Know where you actually sit
Most founders guess high. Run the number properly first, then compare it to the market.
The average digital agency generates around $172,000 per full-time employee, up from roughly $135,000 in 2015. Split by type, the picture is less flattering: marketing agencies averaged about $163,000 per FTE in 2025, blended agencies $167,000, and development agencies just $120,000. Specialists — the ones with a narrow, well-known offer — routinely clear $250,000.
| Revenue per FTE | What it usually means |
|---|---|
| Under $120k | Structural. Overstaffed, underpriced, or drowning in non-billable work |
| $120k–$160k | Surviving. Margins too thin to absorb a lost client |
| $160k–$220k | Where well-run agencies sit |
| $220k–$300k | Real leverage. Usually a specialist offer plus tight delivery |
| $300k+ | Elite, and almost always productised or heavily systemised |
Development shops read low partly because delivery is genuinely labour-heavy, and partly because a lot of them are still pricing time. That's a choice, not a law of physics.
Two rules before you use the number. Count contractors as FTE fractions, or you'll flatter yourself by outsourcing the denominator. And if you're a marketing agency passing through media spend, use gross income, not billings — otherwise the metric measures your clients' ad budgets, not your business.
2. Read it against margin, not on its own
Revenue per employee is a leverage signal, not a profit measure. The two only tell the truth together.
The average digital agency earned a 13% after-tax net margin in 2025, below the long-run average of about 15% since 2015. What's striking is the shape of it: studios under 10 people averaged 19%, while firms with 50+ people averaged 8%. Scale, for most agencies, has been actively destroying margin.
That's the trap the founder above walked into. He grew the thing that shows up in the pitch deck and quietly diluted the thing that shows up in his account. High revenue per employee with thin margins means you're pricing well and spending badly. Decent margins with low revenue per employee means you're under-earning per head and holding profit together through frugality — which works right up to the first bad quarter.
Track both quarterly. On the same slide.
3. Attack the denominator before you touch the numerator
Every founder's instinct when this number looks bad is to raise prices. Do that, but understand it's the slower lever, because it's gated by the market and your sales cycle.
The denominator moves faster. The average agency spends 25–35% of total available hours on non-billable activity. Industry utilisation sits around 55–60%, against a sensible target of 65–80% for producing roles. That gap is not laziness. It's status meetings, chasing approvals, rewriting the same proposal, hunting for the latest version of a file, and reconstructing what a client said three weeks ago.
Take a 15-person agency at $2.4m — $160,000 per head. Recover eight billable hours per person per week from admin drag, and you've added roughly 6,000 hours a year of capacity without a single hire. Even half-billed at $120, that's $360,000, which lifts you to $184,000 per head. Same team, same clients.
Before you post a job ad, work out how many hours you'd get back by fixing your delivery process instead. The hire costs $80k plus overhead. The process fix usually costs a fortnight of discipline.
4. Use AI where it changes the ratio, not where it looks good
This is the year the leverage argument became concrete. Stanford HAI's meta-analysis puts measured productivity gains at roughly +14–15% in customer support, +26% in software development and up to +50% in marketing output where AI is genuinely deployed.
Deployed is the load-bearing word. Survey data on agentic adoption shows a large cohort — around 22% — with negative ROI, and organisations with a named owner for their AI systems convert to production at about 2.7x the rate of those without. Tools bought by everyone and owned by no one make the ratio worse, because you've added licence cost to the numerator's denominator and changed nothing about how work gets done.
The honest test: name the repeatable task, measure the hours it takes today, automate it, measure again. If the hours don't move, the tool is theatre. I've written more about scoping this kind of work properly in how to scope AI automation projects, and it applies to your own operations as much as it does to client work. If you sell leverage to clients, you should be able to show it in your own ratio first.
5. Fix your seniority mix
Revenue per employee is often a hiring-shape problem wearing a pricing costume.
Agencies default to hiring juniors because they're cheap. Then the seniors spend their week reviewing junior work instead of billing, and utilisation collapses at the exact end of the business where the rate is highest. You've lowered your average cost per head and lowered your revenue per head further.
The counter-move isn't "hire only seniors". It's to be deliberate about the ratio. Every junior you add needs a defined supervision cost in senior hours, and that cost has to be priced into the work or absorbed knowingly as an investment. If you can't say how many senior hours a new junior will consume in month one, you're not hiring — you're hoping.
Same logic applies to the middle. Overhead should sit at roughly 20–30% of agency gross income. Above 35–40% and non-billable roles are eating the ratio faster than delivery can feed it.
6. Stop selling hours if you want the number to move
There's a ceiling built into time-based pricing. A person has about 1,800 working hours a year, of which maybe 1,300 are billable in a good year. At $120 an hour, that's $156,000 — and you can now see exactly why the industry average sits where it does. Time and materials caps your revenue per employee at your rate times your utilisation, forever.
Productised and fixed-scope work breaks that link, because efficiency gains accrue to you rather than being handed back as a smaller invoice. The tenth time you deliver the same well-defined engagement, it takes 40% less effort and earns the same fee. That's the entire mechanism behind specialist agencies hitting $250k+ per head: not higher rates, but the same fee against falling delivery cost.
The trade is real risk transfer, which is why the choice deserves proper thought rather than ideology — I've laid out both sides in fixed price vs time and materials for agencies.
7. Make it a decision rule, not a dashboard number
A metric you only look at is a metric that changes nothing. Turn it into a rule your team can apply without you.
Set a floor — say $180,000 per FTE — and make it a gate on hiring. Any new headcount must come with a stated path to clearing the floor within two quarters, in writing. If the answer is "we're just swamped", the honest translation is usually that you're underpriced, over-scoped, or leaking hours to admin.
Review it quarterly against three inputs: average project value, utilisation, and headcount. One of those three explains any movement. Then look at where the drag actually accumulates — most agencies find it in approval cycles and rework rather than in the delivery itself, which is the case I made in why slow client approvals hurt agency cash flow.
This is the part that needs to be structural rather than heroic. If milestones, review requests and sign-offs live in someone's inbox, nobody can tell you where the hours went. We built Scopeyard so that approvals and delivery milestones sit in one place with dates attached — mostly because I got tired of reconstructing that history by hand every quarter. Whatever you use, the requirement is the same: the drag has to be visible before it can be priced out. Scopeyard for product agencies is where we've written up how we think about that.
Final thoughts
I've never met a founder who regretted running this number. I've met plenty who put it off, because they suspected what it would say and preferred the version of the story where revenue was up 40%.
Revenue per employee doesn't care about your growth narrative. It just asks whether adding people made the business better or bigger. Those are not the same thing, and only one of them pays you.
Growth that doesn't raise revenue per head isn't scale. It's just a larger payroll with the same profit.