When to Fire a Client (And the Math That Makes It Obvious)

Most agency owners hold on too long. The math feels safe — existing revenue beats the uncertainty of finding someone new. But a bad client doesn't just cost you money. They cost you your best people and your next good client.

Here's how to know when to pull the trigger.

The Warning Signs That Actually Matter

Not every difficult client is a fireable offense. Demanding clients can be profitable. Disorganized clients can improve. What you're looking for is a pattern of behaviors that structurally damage your business.

They consistently pay late. Net-60 creeping to Net-90 with excuses every cycle isn't a cash flow inconvenience — it's a loan you never agreed to give. If you're floating more than 15% of monthly revenue for a single client, you have a lender-borrower relationship, not an agency-client one.

They've eroded your margins past the point of recovery. Scope creep compounds. A client who started at 40% margin is now at 8% after six months of "quick favors." Track your actual hours. Most agencies discover at least one client where fully-loaded cost exceeds retainer by 20–30%.

Your team is quietly dreading their account. When your senior strategist asks to be moved off an account, listen. Turnover from toxic clients costs real money — replacing a mid-level hire runs $15,000–$25,000 when you factor in recruiting, onboarding, and lost productivity.

They're making your work worse. Some clients override every recommendation, ignore data, and blame the results. They're not a client anymore — they're a case study in what you'd never show a prospect.

The Math You're Avoiding

Here's a calculation most agency operators don't run:

Take one problem client paying $8,000/month. They require 3x the account management hours of a comparable client, they've dragged two team members into performance conversations, and your close rate on new business has slipped because you're too stretched to pitch properly.

That $96K annual contract may actually be costing you $120K+ when you account for real time, team morale impact, and opportunity cost. The client you couldn't take because you had no capacity was offering $12,000/month.

The question isn't whether you can afford to lose them. It's whether you can afford to keep them.

How to Actually Do It

Don't ghost, don't manufacture a conflict, and don't let it drag out over a quarter of passive-aggressive service delivery.

Set a firm offboarding timeline. Thirty days is standard. Sixty if the account is genuinely complex. Put it in writing.

Give them a real handoff. Document their campaigns, hand over assets, make the transition clean. You'll want the referral someday, and the industry is smaller than it looks.

Be direct about the reason — briefly. You don't owe them a forensic breakdown, but "we're refocusing on clients in X category" or "our capacity is shifting" closes the conversation faster than vague language that invites negotiation.

Don't renegotiate at the door. They will offer more money. Sometimes that's worth taking. More often, it buys you another six months of the same problems at a slightly higher rate.

One Exception Worth Naming

If a client is difficult but the account is genuinely strategic — a logos client, a vertical you're building toward, a relationship with future acquisition value — that calculus is different. Difficult and valuable isn't the same as difficult and unprofitable. Know which one you have.

What Comes After

Agencies that have fired a bad client describe the same thing: within 60–90 days, that capacity fills with better work. Not always immediately. But the bandwidth, the morale lift, and the sharpened focus almost always compound into something worth more than what you walked away from.

Protecting your team's energy is a revenue strategy, not a feelings strategy.

You built the agency to do good work at scale. Clients who make that impossible aren't just annoying — they're a structural threat.

---

Action Items

• Audit your current client roster. Score each on margin, hours consumed, and team sentiment. The problem account usually surfaces immediately.
• Set a margin floor. Below 25% fully-loaded, a client needs a plan to fix it or an exit date.
• Build the off-ramp now. Draft a standard offboarding process before you need it. Emotional decisions made under pressure are almost always worse.
• Track morale as a metric. Ask your team leads quarterly which accounts they'd cut first. They already know the answer.