How to Know When to Fire a Client (and When Not To)

Every owner we've worked with in the past twenty years has had one client who pays on time and still makes the whole week worse. The invoice looks fine. The account is current. But the calls run long, the scope keeps creeping, and somewhere around month four you start dreading their name in your inbox. That's usually the moment an owner starts asking us when to fire a client — not because the money is bad, but because the math underneath the money has quietly gone bad.

Most advice on this treats it as a personality question: is the client rude, demanding, unreasonable. That's the wrong test. A demanding client who pays a fair price for the extra work is a fine client. The one worth firing is the one whose real cost, once you do the math, is higher than what they pay. Here's the math we walk owners through, and what we tell them to do about it this week.

The client costing more than the invoice shows

Take a contractor billing $85 an hour. One client calls twice a week to re-explain a change order that was already agreed on, then disputes the invoice when the work matches what was discussed. Six unpaid hours a month of calls, redrafts, and follow-up isn't unusual for an account like that. At $85 an hour, that's $510 a month given away for free — over $6,000 a year, before you count the stress of the account.

A bakery doing $40k a month in wholesale can have the same problem dressed up differently. One café account orders inconsistently, pays net-45 instead of the net-15 everyone else agreed to, and forces two extra van reroutes a week to hit their pickup window. The revenue line looks healthy. The driver's schedule and the cash flow do not.

When to fire a client: the math we use

We ask owners to run one number: take what a client pays in a month and divide it by every hour spent on that account, not just the billable ones. Include the extra calls, the redone work, the emails chasing payment, the meetings that ran long because the client wasn't prepared. That's the account's real effective rate.

Compare that number to your normal rate. If it's close, the client is fine, even if they're a handful. If it's meaningfully below — say a dental practice's front desk spends four extra hours a month chasing one patient family's insurance paperwork and re-explaining billing that was already explained twice — that account is quietly subsidized by every other client on the books.

A boutique sees the same pattern with a customer who returns half of what she buys, always past the posted window, and expects the staff to make an exception every time. The transactions look like sales. The time spent processing them isn't free, and it isn't reflected anywhere in the register total.

Owners resist this math because firing a client feels like losing revenue, and revenue feels safer than it is. But the $510 a month a contractor gives away in unpaid calls, or the two extra van trips a bakery makes for one café, isn't safety. It's capacity that could go toward a client who pays for the hours they actually use, or toward the owner's own time back.

Three signs the relationship is already over

When the math confirms what an owner already suspects, we look for three patterns before recommending anything final. First, a payment pattern that's genuinely different from the rest of the client list — not one late invoice, but a repeated habit of disputing or delaying. Second, scope creep with no consequence: the client keeps asking for more and never adjusts the price or the deadline when you say yes.

Third, and the one owners underrate most: the team dreads the account. A service company with a van and four techs can absorb one difficult stop on the schedule. When the dispatcher starts quietly rearranging routes to avoid a specific address, that's a cost too, even though it never shows up on an invoice.

Not every hard client fails all three tests, and that matters. A dental office's most demanding patient family might pay promptly, follow every recommendation, and simply need more reassurance than most. That's a client worth keeping and managing differently, not firing. The math is there to separate the two, not to justify getting rid of anyone who's difficult.

In practice: how USBN works through this with clients

When this comes up in a consulting engagement, the team behind USBN doesn't start with opinions about the client. It starts with a short account-level cost audit: hours logged against that account for the last three months, matched against what it actually paid. That turns a gut feeling into a number an owner can act on instead of an argument they can't win with themselves.

From there the conversation usually splits into two buckets: accounts that are hard but genuinely profitable once you count everything, and accounts that are hard and quietly losing money. The first group gets boundaries — clearer scope, a kill fee for redone work, a faster path to a price conversation. The second group gets a plan to wind down: a fixed end date, a referral to someone better suited, and a conversation that stays professional, because a Brooklyn business market is smaller than it looks and today's dismissed client is tomorrow's reference check. If the real issue turns out to be that your prices across the board haven't kept up, that's a related conversation worth having; see our piece on <a href="/insights/when-to-raise-your-prices">when to raise your small business prices</a>.

What to do this week

Pick your three most draining accounts. For each one, write down what they paid last month and every hour anyone on your team spent on them — calls, redone work, chasing payment, the works. Divide one by the other. If a number comes back well below what you'd normally charge for that hour, you have your answer, and it isn't a personality question anymore.

This kind of account-level math is part of what USBN's <a href="/services#consulting">consulting work</a> covers for growing businesses. If the number surprises you and you want a second set of eyes on it, that's exactly what a <a href="/contact">strategy call</a> with us is for.