Client Profitability: The Math Behind Keeping or Firing a Client

The number on the invoice is not your rate. Your rate is what you keep per hour of your life.

A client paying $5,000 a month looks good until you add up the hours: the scope creep, the re-explanation calls, the Sunday Slack messages, the invoice follow-ups, the revision cycles that weren't in the original agreement. At 80 hours a month, that's $62.50 an hour. A client paying $2,000 a month who needs 10 hours of your time is $200 an hour. By revenue the first client wins. By every metric that actually matters, they don't.

Effective hourly rate is the number to track. Everything else is misleading.

How to calculate it

Effective hourly rate = total annual revenue from the client / total hours worked for that client annually.

Count all the hours. The monthly review call. The two rounds of revisions that weren't scoped. The time you spent answering their questions about things outside the engagement. The admin time chasing payment. If you touched it, it counts.

Most freelancers track revenue per client precisely and hours per client not at all. The ones who start tracking hours are usually surprised by which clients actually pay the best rate — and which ones they've been subsidizing.

When the math says something is wrong

Three situations where the numbers usually justify action:

The client's effective rate is below your floor. If you've set a rate you're willing to work for — say $125 an hour — and a client is paying $80, they're being subsidized by the margin on your other clients. You're doing below-floor work and it's filling capacity that a better client could occupy.

They're over 30% of your revenue. Concentration risk compounds the rate problem. One client at 30% of income means your livelihood depends on a single relationship continuing. The rate math and the dependency problem usually appear together.

They've held you from better work. You've declined projects or stopped pursuing new clients because this client fills your calendar. The opportunity cost is real even when the effective rate looks acceptable. The question isn't just what they're paying you — it's what you're not getting paid because they're there.

What the math looks like in practice

A client pays $3,000 a month and takes 30 hours. Effective rate: $100 an hour.

Your target is $150 an hour. At 30 hours a month, that slot is worth $4,500 at your target rate.

The gap is $1,500 a month, or $18,000 a year. That's the annual cost of keeping one below-rate client when the slot could hold a better one.

Whether you can close that gap depends on your pipeline. If you have demand at your target rate, the math is straightforward. If your pipeline is thin, the problem is pipeline, not whether to eventually fire the client.

How to exit without making it worse

Give 30 days notice for most engagements. Give 60 for long-running complex work where the client's team would need meaningful time to transition.

Frame it as a capacity shift, not a judgment. "I'm narrowing my focus and won't be able to serve you as well as you deserve going forward" is neutral, professional, and true. Most clients accept it without issue.

Don't explain the rate math. They didn't hire you to advise on your business decisions.

If they ask to negotiate, that conversation sometimes reveals what they'd actually pay to keep you. A number that changes the math warrants a genuine conversation. A number that doesn't — one that's still below your floor — confirms you made the right call.

The sequencing mistake

The most common error is firing a client before the replacement revenue is in place.

A 30-day pipeline cycle means you need a prospect at the offer stage before you give notice. If you don't have that, the principled business decision becomes a cash flow problem. Start the replacement process first. Fire when you have something to land on.

This sounds obvious. It is not how most people do it.

The Client Profitability Calculator ranks all your clients by effective hourly rate, shows you your blended rate across the full book, and makes visible exactly which clients are pulling it down.

Results are based on inputs you provide. Hours estimates should reflect actual averages over a representative period — not best-case or estimate-from-memory figures.