Should I fire a client, or fix the deal?

Everybody says try to fix it first. Nobody says why you won't: a price increase can be refused and a termination can't, so firing feels like the easier one.

When somebody asks me this, they already have a name in their head.

It is almost always the wrong name.

The client you are picturing is the rude one. Saturday texts, three-week approval cycles, feedback that arrives as a phone call at six. That client is exhausting, and in my experience that client is usually profitable, because the friction is visible and visible friction gets priced into the next renewal.

The one worth looking at is the account everyone on your team describes as solid. Pays on time, no drama, been there three years. Nobody would nominate it. It is also, quietly, a different piece of work from the one you sell now, at a price you agreed when you were worse at this.

So here’s my answer, and it isn’t the brave one. Most of the time you shouldn’t fire the client. You should reprice or rescope the work. And the reason people skip that step and go straight to firing is that a price increase can be refused and a termination can’t.

Firing is the conflict-averse option wearing the brave option’s clothes.

Three questions that sort it in ten minutes

What is this account’s real hourly rate? Take last quarter’s fees from them and divide by every hour that actually went in. Not the hours you billed. The calls, the revisions, the Slack channel somebody monitors, the internal escalation about their internal escalation. I’m not going to hand you a benchmark here because the honest ones don’t exist, but I’ve never seen an owner do this calculation and find the number was better than they thought.

Would you take this work today, at this price? If the answer is no, you don’t have a client problem. You have a pricing problem that predates them and is probably also in your two newest contracts.

Does the work they ask for match what you want to be hired for in two years? The account that pays fine and teaches you nothing is more expensive than it looks, because it occupies the capacity you would need to build the thing you actually want to sell.

If your answer to any of this starts with “but we can’t afford to lose them”, stop. That’s a concentration question rather than a firing one, and I went through how much revenue from one client is too much separately.

Everybody says fix it first. Nobody says why you won’t

The advice to try fixing the relationship before ending it is close to universal, and it’s correct. Karl Sakas, who advises agencies on exactly this, puts the sharpest version of it plainly: if your agency is frustrated with a client, they’re probably frustrated with your agency too.

Fine. So why does nobody do it?

Because the repricing conversation can be turned down. You ask for forty percent more, they say no, and now you’re sitting in a relationship you’ve just told them you resent, at the old price, with less leverage than you had last week. That outcome is genuinely worse than today.

A termination cannot be turned down. It’s the only conversation in the whole situation where you control the ending. That is why people take it, and almost nobody admits that’s the reason.

The second reason is that fixing it requires you to still be open to staying, and by the time somebody asks me this question they have usually decided. The conversation they then have is not a renegotiation. It’s a notice period with extra steps.

If you do want to fix it, the thing people leave out is the consequence. Name the specific behavior, state the new price or the narrower scope, give a date, and say what happens if nothing changes. The last clause is the entire mechanism. Without it you’ve made a complaint.

The one profession that had to write this down

Marketing has opinions about this. Law has rules.

Washington’s rule of professional conduct 1.16, in the version amended effective 1 September 2026, says a lawyer may withdraw where the representation “will result in an unreasonable financial burden on the lawyer or has been rendered unreasonably difficult by the client.” So “this account is not worth what it costs me” is a recognized ground for leaving, written down, in a rule enforced by a disciplinary board. You are allowed to go.

But read what sits next to it. Where the problem is that the client has failed to meet an obligation to the lawyer, withdrawal is permitted where the client “has been given reasonable warning that the lawyer will withdraw unless the obligation is fulfilled.”

The warning is in the rule.

None of this binds you. You’re not a lawyer and a marketing retainer is not a representation. It’s worth knowing about because this is the one profession that had to decide the question under pressure, in writing, with a regulator watching, and what it landed on was not a list of red flags. It was a requirement to tell the client first.

The exit is where it actually costs you

The same rule has a part that gets skipped, and it’s the part I’d copy. On termination, a lawyer has to take reasonable steps to protect the client’s interests, “such as giving reasonable notice to the client, allowing time for employment of another legal practitioner, surrendering papers and property to which the client is entitled and refunding any advance payment of fee or expense that has not been earned or incurred.”

Three obligations, and all three translate.

Notice long enough to be useful. Not your contract’s thirty days if thirty days leaves them mid-campaign with nobody hired. Notice is measured in whether they can replace you, not in whether you complied.

Hand over the assets, properly. Ad account access with admin rights, the GA4 property and Search Console, the domain and the DNS, CMS logins, source files, and the documentation of what is tracking what. The one that causes the fight is always the same: accounts created inside your own manager account, which feel like yours and are not. Sort that out while you still like each other.

Refund what you haven’t earned. Prepaid work you won’t do goes back. Arguing about the last two thousand dollars is how a quiet exit becomes a review.

Then don’t badmouth them. The market is smaller than it looks and the person you are relieved to lose will be somebody’s VP in two years.

Where I come out

Firing a client you cannot replace is a luxury decision, and most of the writing about it comes from people with a full pipeline. If this account is a third of your revenue, you don’t have a firing decision in front of you, you have a sequence: reprice, replace, then leave. Doing it in the other order is not courage, it’s just expensive.

There is one exception, and it’s the only case where I’d skip the repricing conversation entirely. The same lawyers’ rule allows withdrawal when a client insists on action the lawyer finds repugnant or fundamentally disagrees with. In our world that’s the client who wants the review they didn’t earn, the claim you’d have to stop believing to make, the number reported a way you know is wrong.

There’s no price that fixes that one. Go.

If you’re weighing one of these up and want a second read on whether it’s a pricing problem or an ending, drop me a line on email, WhatsApp, phone or LinkedIn and we can have a quick chat, and my notes on how I work cover the rest. I’m taking on a small number of contracts at the moment, so I’ll tell you straight whether it’s something I could help with.

Say hello

Have a number that is not moving?

I am accepting contracts to help brands grow, and I am equally happy with a conversation that never becomes one. Email, LinkedIn, WhatsApp or a call, whichever is easiest. Tell me what you are running and what it is failing to produce, and you will get a straight answer on whether I am the right person for it.

Or call +91 70199 90776.