Terms of Disengagement: Fixing Client Stop Conditions Before You Need Them
Most owner-managed firms lose money on clients in slow motion. The invoices are paid, the work is delivered, and the accounts show a surplus. Yet the partner time is heavier than it should be, the project manager has started to flinch at the name on the email, and the scope has drifted so far that nobody can say with confidence what was originally agreed.
By the time the relationship is discussed properly, the decision has usually been made by fatigue rather than judgement. The purpose of a stop rule is to move that decision earlier — to a point where you still have the capacity to negotiate it.
Why the decision is normally taken too late
An owner-manager facing a difficult but profitable client is in an awkward position. Ending the relationship means accepting a hole in revenue, explaining it to the team and, often, admitting that a decision made some years ago no longer holds. The natural response is to wait and see. Waiting is also the cheapest option in the short term, which is precisely why it persists.
The difficulty is that the moment you most want to exit is the moment you have least energy to do it well. A stop rule replaces a series of small resentments with a pre-agreed test. It does not remove the awkward conversation; it simply ensures the conversation happens before the relationship has become corrosive.
Four thresholds worth committing to paper
A useful stop rule is short enough to remember and specific enough to be applied by someone other than its author. Four thresholds do most of the work.
1. Margin after the true cost of serving
Gross margin flatters difficult clients. The more honest measure is contribution after partner time, project management, rework, escalation meetings and the working capital tied up in slow payment. A practical threshold might be contribution below a set percentage for two consecutive quarters, measured on a rolling twelve-month basis rather than a single invoice. The rolling view prevents a good month from disguising a bad year.
2. Conduct and payment discipline
Behaviour is a legitimate commercial threshold. Persistent late payment, rudeness to staff, disregard for agreed boundaries and a casual relationship with the truth about scope are all measurable over time. The test should distinguish a bad week from a pattern: for example, two written escalations within a rolling six months, or payment beyond terms on three consecutive invoices without prior agreement. A client may be demanding and still worth keeping. A client who is dismissive of your team is a different proposition.
3. Scope drift
Scope creep is rarely dramatic. It arrives as a quick call, a small addition, a favour. The threshold is the point at which unbilled variation exceeds an agreed proportion of fees, or the written scope has not been reviewed for a defined period. If the scope document is more than a quarter out of date, that is itself a signal that the engagement is being run on memory rather than agreement.
4. Reputational cost
The hardest threshold to quantify and usually the most important. It covers work that misrepresents your standards, clients whose names you would rather not cite, and engagements that pull the firm into a sector or practice it does not wish to be known for. A blunt test serves well: would you be content for this engagement to be described, accurately, in front of your peers? If the answer is no, the threshold has already been crossed.
A one-page stop schedule
The rule should fit on a single page and be signed off by whoever holds commercial authority in the business. A simple format is enough.
| Test | Threshold | Trigger | Owner | Action |
|---|---|---|---|---|
| Margin | Contribution below agreed floor for two quarters | Quarterly review | Finance lead | Review and reprice or conclude |
| Conduct | Two written escalations in six months | Any escalation | Account lead | Formal conversation, then review |
| Scope | Unbilled variation above agreed percentage | Monthly | Project lead | Change control or re-scope |
| Reputation | Peer-description test failed | Any time | Principal | Immediate review |
The exact figures matter less than the fact that they are written down, dated and agreed before the relationship becomes difficult. A rule invented after the event is not a rule; it is a justification.
The review routine that makes it work
The rule only functions if it is reviewed on a rhythm. Three mechanisms are sufficient for most firms.
First, a quarterly review of the client base, held before the sales pipeline discussion rather than after it. Forty-five minutes with a standing agenda is enough. Second, a formal review at least ninety days before any renewal or contract anniversary, with a written note of the decision. Third, an event-driven trigger: any single serious breach — a payment failure, an escalation, a reputational concern — calls a review within ten working days.
Each review should end with one of three recorded outcomes: continue, correct, or conclude. A decision to correct is only meaningful if it names the actions, the owner and the date by which the position will be reassessed. Without those three details, correction becomes postponement.
Holding the line when the relationship is pleasant
The awkward part of a stop rule is that it will sometimes point towards ending a relationship you quite like. That is the test of whether it is a governance instrument or a piece of theatre. The rule is not a punishment for a client who has been difficult; it is a prior commitment made by people who knew that future judgement would be clouded.
It is equally important not to use the rule as a substitute for a direct conversation. In most cases the correct first response is to raise the issue plainly, agree a change and set a date to review. The stop rule is the backstop, not the opening move. A client who responds well to a candid conversation is usually worth another year. A client who does not has told you something useful.
What happens at the stop point
When the threshold is met and the decision is to conclude, treat the exit as a project. Check notice periods and termination clauses. Agree a handover. Communicate early to the people who need to know, including your own team. Manage the debtor position before the relationship cools. Redeploy the capacity deliberately rather than letting it drift.
And leave cleanly. A well-managed exit protects the reference, the reputation and the possibility of a future relationship on better terms. The point of a stop rule is not to lose clients. It is to make retention a decision rather than a default — and to ensure that the decision is taken by judgement while there is still time to act on it.
