When Should an Agency Fire an Unprofitable Client? The Margin Math of Letting Go
Nobody starts an agency dreaming of firing clients. But every agency owner has the one: the account that pays on time, looks fine on the revenue line, and quietly eats the margin of everything around it. Firing an unprofitable agency client is not an emotional decision, or at least it should not be. It is a math problem with three possible answers, and most owners run the math years later than they should.
The question is not "how much do they pay?"
A client paying $100,000 a year can be the most expensive relationship in your portfolio. One consultant framed the real test as a single question: "Would I actively pursue this client again if I had the choice today?" If the answer is no, you are keeping a client your current self would reject, which is another way of saying the relationship costs more than it pays.
The accounting version is customer profitability analysis: revenue minus the direct cost of servicing the client, measured honestly. That means the account manager's hours, the revision cycles, the emergency calls, the payment-chasing. An Entrepreneur piece on the subject recommends reviewing the bottom 10 percent of clients by profitability every year, because in most service firms roughly 80 percent of the profit comes from 20 percent of clients, and the administrative headaches cluster at the bottom.
There is a sharper version of the same point from a profitability-modeling paper: a client who is breaking even on paper is mathematically losing you money once the opportunity cost of your team's hours is applied. Those hours could have gone to a client paying your target margin. Break-even is not neutral. It is a loss wearing a disguise.
The three exits, from softest to hardest
Once a client scores low on both margin and qualitative fit, the same analysis lays out three exits:
1. Price them out. Raise their rates until the margin meets your threshold. Some call this the "PITA tax." It is the best first move because it has two good outcomes: either the client becomes profitable, or they leave on their own and you never had to fire anyone. Frame it as a pricing change with a clear effective date, not a personal decision.
2. Refer them downmarket. A client who is unprofitable for you may be a perfect fit for a smaller or cheaper provider. A warm referral costs you nothing, preserves the relationship, and often gets you referred back when the client grows into your price point. This is the exit most agencies skip and most agencies should use first.
3. Terminate outright. End the contract on its terms, with a professional handoff and a transition period. This is for the clients where pricing and referrals are not the answer: the ones who treat your staff badly, ignore your advice, pay late every cycle, or sit outside everything your agency actually does.
IT business consultant Richard Tubb, quoted on the MSP version of this problem, puts it bluntly: agencies mistake cash flow for profit. The big client everyone is afraid to lose often turns out to be not profitable once someone actually runs the numbers.
When to troubleshoot before you exit
MSP consultant Manuel Palachuk recommends scoring clients on profitability plus behavior, payment history, whether they take advice, how they treat your team, and then troubleshooting the low scores before throwing the client away. This is the part the "fire everyone" crowd skips. A client with good margin but bad communication might just need a different account lead. A client with bad margin but great behavior might be one scope conversation away from profitability. The exit is for the client that scores badly on both axes after you have tried the fix.
The related trap: do not confuse a slow quarter with a bad client. One bad project does not make a bad relationship. The firing decision needs a rolling view, two or three projects at minimum, because project margin on a single engagement can be noise. This site has benchmarks for what a good project margin looks like and how scope creep erodes it, which are the inputs to this decision.
Frequently Asked Questions
When should an agency fire an unprofitable client?
When the client is consistently below your margin threshold across multiple projects and scores poorly on behavior too: late payment, constant scope creep, poor communication, or disrespecting your team. Troubleshoot the fixable cases first, reprice the marginal ones, and exit the rest.
How do you know if a client is unprofitable?
Run a customer profitability analysis: the client's revenue minus the direct cost of servicing them, including all unbilled hours. A client breaking even on paper is losing money once opportunity cost is counted. The blunt test: would you pursue this client again if you had the choice today?
Will firing clients hurt agency revenue?
Usually the opposite. Freed capacity goes to better clients, and the price increase offered to remaining clients often more than compensates. The revenue that disappears was low-margin or negative-margin revenue, which was not helping.
What is the gentlest way to fire an agency client?
Raise their price to your profitable rate first. Either they accept and become a good client, or they leave voluntarily. The next option is a warm referral to a cheaper provider. Outright termination with a transition period is the last resort.
Want more of these? I write one practical guide a week on money, careers, and tools that pay for themselves. Subscribe free here.
Know Your True Project Margin
Enter revenue and costs, including the unbilled hours, to see what each client really earns you.
Calculate Project Margin