Your agency runs a 25% blended margin. That number feels fine. It is also the most dangerous figure on your P&L, because client profitability analysis almost always reveals the average is a lie. Some clients earn you 45%. Some earn 5%. A few lose money every month you keep them. The blend hides all of it.
This is how you break profitability down account by account, so you stop guessing and start acting.
The blended margin is an average, and averages lie
A 25% blended margin can mean a dozen different realities. The one that should worry you: three great accounts carrying two that lose money.
Say you run five clients. Two earn 45%. One earns 15%. Two lose 10%. Weight those by revenue and you can still land near 25% blended. The number looks healthy. Two of your five accounts are quietly draining the firm.
This is the distribution problem. Harvard Business Review's work on unprofitable customers makes the point plainly: a small share of accounts often destroys value the rest of the book worked to create. You cannot see that from the top line.
Agency margins compress fast. Promethean Research's 2025 industry report tracks how thin the median firm runs. When the whole book is tight, one loss-making account is the difference between a good year and a scramble.
Why the average points you at the wrong fix
Here is the trap. The blended number pushes owners toward efficiency fixes when the real problem is a few unprofitable accounts.
You see 25% and think: we need faster delivery, cheaper tools, tighter processes. So you squeeze the whole team. You buy new software. You run a process audit across every account, including the 45% ones that were never the problem.
Meanwhile the two loss-making clients keep losing money. You optimized the wrong thing.
HBR's follow-up on profit-drain customers frames the choice better: you fix the account, reprice it, or exit it. None of those is an efficiency project. They are decisions about specific clients — and you can only make them if you can see each client on its own.
How to run client profitability analysis, account by account
You need three inputs per client: revenue, direct cost, and a fair share of overhead. Parakeeto's five-step process is the cleanest walkthrough, and it maps to how agencies actually track work.
Start with the simple version:
- Revenue. What the client paid this period. Straightforward.
- Direct labor cost. Fully loaded hours your team spent on that account. This is where most agencies undercount — you log billable hours but not the strategy calls, the revision rounds, the account management time.
- Contribution margin. Revenue minus direct cost. This is the number that matters most per client. Our contribution margin glossary entry explains why: it shows what each account contributes before shared overhead.
- Allocated overhead. Rent, tools, admin, leadership — spread across accounts by revenue or hours.
- Net margin per client. Contribution minus allocated overhead. This is your net profit margin at the account level.
Run every client through this. Now the distribution is visible. You can see the 45% clients and the loss-makers side by side. NetSuite's guide to agency margins covers the cost categories to include so nothing hides in "general overhead."
If you want to sanity-check contribution math on a single account, our margin calculator does it in under a minute.
The costs that turn a 15% client into a loss
Some accounts look profitable on paper and lose money in practice. The gap is almost always unbilled time and acquisition cost.
Scope creep is the usual culprit. The client asks for "one more round," a quick deck, a Friday call. None gets invoiced. Over a quarter, that unbilled labor erases the margin. We break this down in our scope creep hub — it is the single most common reason a healthy retainer goes underwater.
Then there is acquisition. If you spent heavily to land a client and they churn in eight months, the customer acquisition cost never got recovered. That account was a loss the day it signed.
Promethean's financial metrics guide lists the ratios worth watching here — utilization, delivery margin, and the gap between quoted and actual hours. Track them per client and the leaks show up early. Our agency KPIs hub covers which ones to put on the dashboard.
What to do with each account once you can see it
Analysis without action is just a spreadsheet. Once you have net margin per client, sort them into three moves.
Keep and protect. Your 40%-plus accounts. Do nothing clever. Do not "optimize" them into the ground. Protect the relationship and the scope.
Renegotiate. The 5–15% clients that could work at the right price or the right scope. Take the analysis into the conversation. "Here is what we're delivering, here is what it costs." Repricing is easier when you have the number in front of you. Our pricing models hub covers structures that hold margin better than hourly.
Fix or exit. The loss-makers. First ask whether scope discipline can save them. If the work is fundamentally underpriced and the client won't move, exit. Firing a client feels like losing revenue. You are losing money — cutting the account frees capacity for a profitable one.
SCORE's financial management resources are a solid grounding if you want to build this discipline into a monthly rhythm rather than a one-off panic.
Make this a monthly habit, not a rescue mission
The agencies that stay profitable don't run this analysis once a year in a crisis. They see client-level margin every month, so a slipping account gets caught at 12% instead of at negative 8%.
That is the whole case for a real-time view instead of a spreadsheet you rebuild each quarter. CentSight is the intelligence layer on top of QuickBooks and your bank — connected data, synced on demand, so you can watch client margins as they move. More on the full approach in our agency finance pillar and the client profitability hub. For the delivery side of the picture, see project profitability and the deeper playbook in agency profitability secrets.
The takeaway: your blended margin is not your margin. It is a curtain. Pull it back, look at each account, and the two decisions worth making — renegotiate or exit — will be obvious. A fractional CFO to build this costs $8K–$15K a month. The number you need is already sitting in your ledger.
