When should an agency fire an unprofitable client?
Firing clients has become a genre of business advice, which is unfortunate, because replacing revenue costs considerably more than repairing it.
Exit is the fourth option, not the first. Before ending a client relationship, work through repricing, rescoping and re-resourcing — each of which costs far less than replacing the revenue. A useful trigger for the conversation is an account sitting materially below your target gross margin across two or more quarters, but the number matters less than the cause. An account that is unprofitable because it is underpriced is fixable. One that is unprofitable because the relationship is dysfunctional usually is not.
First, check the number is real
A meaningful share of unprofitable clients are a measurement artifact.
Before acting on a client margin figure, confirm three things. That revenue is stated net of any pass-through spend — if media is inside the fee, the margin is understated for the accounts with the largest budgets. That delivery cost is separated from overhead, so you are not allocating the finance team to a client. And that the loaded cost is honest, not a salary-derived hourly rate that ignores taxes, benefits and realistic available hours.
Get any of those wrong and you may be about to fire a profitable client. The three most common calculation errors are here.
Then check the hours themselves. If time is entered on Fridays in round numbers against a project code that also absorbs two other things, the client margin inherits that. Bad time data is the single most common reason an account appears to be losing money.
The four levers, in order
1. Reprice
The cheapest fix and the one owners most avoid. If the account has been on the same fee for three years while scope grew, the conversation is not adversarial — it is overdue, and most clients know it. Bring the hours data. "This engagement now takes 40% more hours than we scoped in 2023" is a fact. "We need more money" is a request.
2. Rescope
If the client cannot or will not pay more, reduce what they get. Fewer deliverables, longer cycles, a lower service tier. This is frequently the outcome the client prefers, because it protects their budget, and it gets you to the same margin from the other direction.
3. Re-resource
If a senior person is doing work a mid-level person could do, the account is expensive because of how you staffed it rather than how you priced it. Moving delivery down a level, or moving repetitive work to a standard process, can close a surprising gap without any client conversation at all.
4. Exit
Only after the first three have been tried or genuinely ruled out. Exiting is the most expensive option because you lose the revenue immediately and pay to replace it later.
Worth doing the arithmetic before you get to option four. If it costs you a meaningful chunk of a year's fee to win and onboard a replacement client, then an account you could have repriced to break-even was worth repairing rather than replacing — even at a mediocre margin.
When it really is time to go
Some accounts should end, and the signals are usually behavioural rather than financial.
- You have repriced once and the same pattern returned. The problem is the working relationship, not the fee.
- Scope changes arrive without any expectation of paying for them, repeatedly, after being addressed.
- The account is costing you people. If good staff are asking not to work on it, the real cost is turnover and it does not show up in the margin.
- Payment is chronically late and you are financing them.
- They are a reference risk. Some relationships damage you in the market more than the revenue is worth.
- The work is drifting outside what you do. An account pulling you into services you do not want to build is a strategic cost, not a margin one.
Notice that only one of those is a number. Sustained low margin is a prompt to investigate, not a verdict.
How to exit well
- Give real notice. Whatever the contract says, more is better. The goal is a former client who speaks well of you.
- Offer a genuine transition. Documentation, handover calls, introductions to someone who fits them better. This costs little and buys a lot.
- Be honest but not brutal. "We're not the right fit for where this is going" is true and sufficient. A detailed account of everything that went wrong helps nobody.
- Finish the work in flight. Walking away mid-campaign is how an exit becomes a reputational problem.
- Collect what you are owed first. Have the receivables conversation before the exit conversation, not after.
Then check your concentration. Losing an account changes what your largest client represents as a share of revenue, and if that number moves above a comfortable threshold, the exit created a different risk than the one it solved.
Common questions
At what margin should you fire an agency client?−
There is no universal threshold. An account materially below your target gross margin across two or more quarters is worth investigating, but the cause matters more than the number. Underpricing is fixable by repricing or rescoping; a dysfunctional relationship usually is not. Verify the margin calculation is correct before acting on it.
Is it better to fire a client or reprice them?+
Reprice first, almost always. Exiting means losing the revenue immediately and paying to replace it later, which typically costs more than the margin you were losing. Repricing, rescoping and re-resourcing are all cheaper, and clients on stale fees frequently expect the conversation.
How do I know which clients are unprofitable?+
Compare fee income per client against delivery cost per client, where delivery cost is hours multiplied by fully loaded cost plus any client-specific direct costs. Revenue must be stated net of pass-through spend, and delivery cost must exclude overhead. Without those two corrections the answer is unreliable.
How much notice should you give when ending a client relationship?+
At least what the contract requires and preferably more. The objective is a former client who refers you rather than warns people about you, and generous notice plus a real handover is the cheapest way to buy that.
What if the unprofitable client is our largest?+
Then repricing matters more and exiting is riskier, because losing them creates a concentration and cash problem on top of the margin one. Work the first three levers harder, and if you do exit, plan the replacement revenue before giving notice rather than after.
See it before you decide
Margin per client, monthly. Most agencies find the problem is repricing, not exiting.
Get a free books reviewGeneral information for agency owners and operators, not accounting, tax or legal advice. Figures shown are illustrative unless stated otherwise.