Should you take on a client who is a bad fit
Separate the four kinds of bad fit so you know which ones can be managed with conditions and which one should always be declined outright.
The deal you can already see going wrong
You know the signs by the second call. The brief keeps changing. The budget does not match the ask. Somebody mentions that the last three agencies did not understand their business. And the invoice value would still be welcome this quarter.
The useful move is not to ask whether the client is difficult. Almost every client is difficult in some direction. The useful move is to identify which kind of bad fit you are looking at because three of the four kinds can be managed with conditions and one of them cannot be managed at all.
What accepting actually costs
A bad fit client does not become a reference and does not teach your team anything you wanted to learn. Strip those two justifications out first because they are the ones owners use to talk themselves into the deal.
The costs are mostly invisible on the invoice. Senior time gets pulled into an account that was priced for junior delivery. The project overruns and the overrun displaces work for clients who pay properly. Collection effort arrives at the end and an account that argues about scope usually argues about the final payment too. There is a morale cost as well. Your best people notice which accounts they are protected from and which they are handed.
When taking it is the right call
A business in its first year with no portfolio and no cash buffer should sometimes take the awkward deal. Survival comes before selectivity and pretending otherwise is advice written by people who already have a pipeline. Take it with conditions rather than with hope.
A business at or near capacity should decline. Every hour a bad fit consumes is an hour removed from an account that pays on time and refers you onward. The business that should always decline is one where the bad fit is payment behaviour. Scope confusion is fixable through documentation. Budget mismatch is fixable by cutting scope. A buyer who has already told you they pay late or who disputes invoices as a matter of habit is telling you the ending in advance.
The four kinds and what to do with each
The wrong problem means they want something you do not do well. Decline and refer. Taking it produces work you cannot show anybody and a client who is correctly unhappy. The wrong budget means they want your full scope at a fraction of the price. That is solvable. Sell a smaller piece at the published rate and state the exclusions in the first line.
The wrong timing means they are not ready. No approved budget. No internal owner. A restructure under way. Put the proposal aside with a date to return rather than pushing it through. The wrong behaviour means they treat suppliers as adversaries. Nothing in your process fixes that and no amount of documentation makes it pleasant.
When you do accept a risky account attach conditions before the first invoice. A smaller first phase with a defined end. Payment in advance for that phase. The scope in countable units with a written definition of done. A named exit point where either side can stop. Conditions are not hostile. They are the only honest way to accept an account you have doubts about.
Whether the decision was right
The metric that matters is margin after overrun combined with how much senior time the account consumed. Both must be read together. A profitable account that absorbed every hour of your most capable person has blocked your ability to grow.
The metric that flatters is revenue added. It records the invoice and ignores everything the invoice cost. Team utilisation flatters in the same way. A fully utilised team working on a loss making account looks identical in the report to a fully utilised team working on your best one.
The mistake that doubles the loss
The common error is taking the bad fit and then discounting it to make the awkwardness feel acceptable. The reasoning sounds sensible. The account is difficult so charge less and expect less. In practice the discount removes the only buffer that made the overrun survivable.
A discounted bad fit account also sets the price for everything that follows. The client refers somebody at that rate. Your team now believes difficult accounts are cheap accounts. If you accept a risky client then accept them at the full rate with a smaller scope. Price is the compensation for the risk and giving it away leaves you carrying the risk for free.
Declining without burning the contact
Say no in one short written message that names the reason and offers the alternative. The reason should be about fit rather than about them. Many declined buyers return eighteen months later with a proper budget and a clearer brief. That only happens when the no was specific and polite rather than vague and slow.
This guide explains how something works. It is not legal, tax or financial advice, and rules change. Confirm anything regulatory with a qualified professional before you act on it.