You are not deciding whether the client is difficult. You are deciding whether this engagement, at this price, with this behaviour, is a better use of a slot than what you would put in it instead.

Signals worth acting on

  • Chronic non-response that has survived a direct conversation about it.
  • An effective rate well below your floor after a repricing attempt has been declined.
  • Boundary violations that continue after being addressed explicitly.
  • Being asked to do something you are not comfortable with — this one is not a capacity question and should not be treated as one.
  • Consistent late or disputed payment.
  • How you feel when their name appears. Unscientific and highly reliable.
One signal that is not enough on its own: a client being demanding. Demanding-and-well-priced is a normal professional relationship. Demanding-and-underpriced is a pricing problem, and worth trying to fix before it becomes a termination.

Try this first

Firing a client costs revenue, replacement effort and sometimes reputation. Two cheaper options usually deserve a try:

  • Reprice. Many "bad clients" are simply underpriced ones. At the right fee, the same behaviour is tolerable — and if they decline, they have made the decision for you, which is easier for both sides.
  • Reduce scope. Sometimes one part of the engagement causes all the friction. Removing it can save the rest.
  • Have the direct conversation once. Not a hint — a specific, calm statement of what needs to change. Some clients genuinely did not know, and change immediately.

If a repricing and one direct conversation have both failed, you have your answer.

Timing

  • Not mid-deadline. Leaving a client stranded before a filing is bad for them and worse for you.
  • Ideally at a natural boundary — after a year end, after a close, at a renewal date.
  • With the notice your engagement letter specifies. If it does not specify one, add it before your next client.
  • Not when you are angry. Sleep on it once. The decision will look the same and the email will read better.

What to say

Brief, unambiguous, no blame. A long explanation invites negotiation and gives them something to argue with.

Subject: Ending our engagement — {{date}}

Hi {{first name}},

I've decided to end our engagement, with effect from {{date}}. That gives {{n}} weeks for a handover.

This isn't about anything you've done wrong — my practice is moving in a direction where I'm not the right fit for your needs, and you'll be better served by someone who is.

Between now and then I'll complete {{work in progress}}, and I'll prepare a full handover pack: {{list}}. I'm happy to speak to whoever takes over.

If it would help, I can suggest a couple of bookkeepers who'd suit your business well.

Thanks for the work we've done together.

{{your name}}

"Not the right fit" is doing deliberate work here. It is true, it is unarguable, and it does not invite a debate about whose fault anything was. Detailed grievances feel satisfying to write and make the next four weeks considerably worse.

Leave it clean

How you exit is what gets talked about. Do the handover properly even when the relationship was poor:

  • Bring the books to a defined, documented point.
  • Provide a full handover pack — file access, reconciliation status, open items, working papers.
  • Remove yourself from software access only after the client has confirmed their own.
  • Settle the final invoice cleanly, and do not withhold records as leverage.
  • Reply to the successor's reasonable questions for a short, stated period.

Bookkeeping is a referral business, and the successor bookkeeper is a person who refers work. A clean handover from a difficult exit is more valuable than most marketing.

Before firing, check whether the friction was collection

A large share of "difficult client" experiences are really document-chasing experiences. Automating the requests, reminders and status tracking often changes how an engagement feels enough to make the question go away.

Start your free week →