Agency profit margins: what good looks like at 5, 15 and 50 people

The assumption underneath most agency growth plans is that margin improves with scale. Hire more people, spread the overhead, get better at delivery, and the percentage goes up.

It doesn't work like that. Margin at an agency is shaped like a valley. Small shops run lean and often make decent money. Large agencies eventually earn the efficiency back. The middle — roughly 12 to 25 people — is where most owners discover they are working harder, billing more, and taking home less.

Here's what that looks like with the numbers held deliberately flat.

Three hypothetical agencies

Same discipline, same market, same revenue per head. Only the size differs. These figures are invented to isolate one variable, not benchmarks pulled from anywhere.

	5 people	15 people	50 people
Revenue	$750,000	$2,250,000	$7,500,000
Revenue per head	$150,000	$150,000	$150,000
Delivery cost	$412,500 (55%)	$1,237,500 (55%)	$3,975,000 (53%)
Gross profit	$337,500 (45%)	$1,012,500 (45%)	$3,525,000 (47%)
Overhead	$232,500 (31%)	$787,500 (35%)	$2,625,000 (35%)
Operating profit	$105,000 (14%)	$225,000 (10%)	$900,000 (12%)

Revenue per head is identical across all three. Gross margin barely moves. The entire swing comes from overhead as a share of revenue, and it goes the wrong way exactly when owners expect it to go the right way.

Note also what happens in absolute terms. The 15-person agency makes $225,000 — more than twice the five-person one — on three times the revenue and three times the risk, headcount and stress. That's the trade a lot of owners make without ever writing it down.

Why the five-person agency does well

Overhead is genuinely small. There is no ops manager, no account director, no second office, no HR system. The owner sells, delivers, and does the invoicing. The cost of coordination is close to zero because everyone is in one conversation.

The catch is that the number is fragile and partly fictional. If the owner is billing 60% of their week and paying themselves below market, the 14% is subsidised by an unpriced salary. Two questions expose it: am I paying myself what I'd have to pay a replacement, and what happens to this number if I stop delivering? At five people, the honest answer to the second question is often "it goes to zero."

Why the fifteen-person agency struggles

This is the squeeze, and it is structural rather than a failure of management.

Between roughly ten and twenty-five people, you buy a management layer before you have the volume to spread it across. In the table above, overhead rises from 31% of revenue to 35% — an extra four points, which is $90,000 a year on $2.25m. That buys perhaps one full-time non-billable role and change.

What forces the spend:

• The owner stops being billable. That's revenue out and no cost out.
• You need someone running delivery. Project managers, resource planning, someone who owns the schedule.
• Admin becomes a job. Payroll, contracts, recruitment, finance, tooling.
• Utilisation drops. More internal meetings, more handovers, more time spent on people rather than work. Every point of utilisation lost shows up directly in delivery cost.

Meanwhile the owner's own selling time collapses, so revenue growth flattens at exactly the moment costs step up.

Why the fifty-person agency recovers

Two things change. Overhead stops growing in steps and starts growing proportionally — one finance hire covers fifteen people or fifty, so the same cost lands on more than three times the revenue. And team mix improves: you can afford juniors and specialists rather than generalist seniors, so the average delivery cost per hour falls even though senior salaries have risen. That's the 53% delivery cost in the table versus 55%.

The recovery isn't automatic. Plenty of fifty-person agencies run at 5%. But the mechanism is available at fifty and simply isn't at fifteen.

How to read your own number

Three definitions matter, and mixing them up is the most common error in this conversation.

Gross margin     = (revenue − delivery cost) / revenue
Operating margin = (revenue − delivery cost − overhead) / revenue
Owner-adjusted   = operating margin with a market-rate salary for every owner

Delivery cost means fully loaded cost for people who deliver client work — not salary alone. If you're using salary divided by 2,080, your gross margin is flattering you by a wide margin.

Overhead is everything else: rent, non-billable salaries, software, insurance, finance, marketing, and the portion of owner time not spent on client work.

The third line is the one to be strict about. If you are taking $60,000 in a market where the role pays $120,000, then $60,000 of what your P&L calls profit is actually a wage you deferred.

What to do with a bad number

If gross margin is weak, the problem is in delivery: pricing, utilisation, team mix, or scope creep. Look at project-level margin before you touch anything else.

If gross margin is fine and operating margin is weak, the problem is overhead. The uncomfortable version of that diagnosis is usually headcount — a role added during a good quarter that never got re-justified.

If you're in the middle band and the number has been falling for two years, you're in the valley. The two ways out are opposite in direction and both work: grow through it fast enough that overhead dilutes, or shrink back to a size where you don't need the layer. What doesn't work is sitting at eighteen people hoping it resolves.

This week

Pull last year's P&L and split every cost line into exactly two buckets: delivery of client work, and everything else. No third bucket, no "partly." Then calculate gross margin and operating margin, and redo operating margin with a market salary for yourself.

Most owners find the split takes an hour and tells them something the annual accounts never did: whether they have a pricing problem or an overhead problem. Those need completely different fixes, and treating one as the other is how agencies spend three years getting nowhere.