The Fire-or-Fix customer audit (Template inside)
Hi Everyone,
When Cassie Young led customer success at the email marketing company Sailthru, her team compared what each customer segment paid against what it cost to serve.
Customers paying under $100K a year brought in 13% of revenue and took up 40% of the customer success team's time. The same segment also had the worst satisfaction scores of any group.
Today we're walking you through the audit that finds this segment in your own customer base, plus a fix that works better than firing anyone.
Why your P&L can't see it
Robert Kaplan at Harvard Business School spent years measuring customer profitability inside companies, and the results repeat almost everywhere.
The most profitable 20% of customers generate between 150% and 300% of total profits. The middle group roughly breaks even. The bottom 10–20% cost more to serve than they pay, and their losses cancel out much of what the top group earns.
Your P&L only shows the net result, so the unprofitable customers keep getting funded year after year. Finance spends time on their billing disputes, engineering builds small custom features for them, and none of it appears under a line called "cost of this customer."
Run the fire-or-fix audit
Open the audit spreadsheet we built and fill out the following:
Revenue share: Group your customers into bands by annual contract size, and work out what percentage of total revenue each band brings in.
Support share: For each band, estimate the percentage of total support and customer success hours it takes up. Support teams usually log their hours, so the data exists once you ask for it.
Satisfaction: Rank the bands by their customer satisfaction scores, from happiest to least happy.
A band whose support share sits far above its revenue share, and whose satisfaction sits at the bottom, is the segment costing you money.
Ask your support or customer success lead this week for hours by account band. Getting that data is the slowest part of the audit; once you have it, the spreadsheet comes together quickly.
Get the fire-or-fix audit sheet here
Check these three places too
Some of the cost never makes it into logged hours. Three more places to check:
Custom work: Ask engineering which accounts their one-off requests come from, and compare that list with what those accounts pay.
Exit interviews: Check whether the same account keeps coming up when people resign or ask to change roles.
Payment and scope: Check which accounts pay late month after month, or keep requesting work outside the contract.
Any account that appears here is worth pricing out individually, even if its segment looks healthy.
Fix before you fire
Sailthru gave its unprofitable segment three options instead of ending the contracts: pay $12K a year to keep a dedicated customer success manager, stay at the current price with basic support only, or move to month-to-month and leave at any time.
Some customers left, and those were the ones costing the most anyway. Most stayed, chose the first option, and the extra $1K a month made the segment profitable again.
When you work through your own list, re-price before you cancel. A segment is usually unprofitable because the price never covered the service, and canceling contracts leaves that pricing mistake in place for the next deals you sign.
Won't they bad-mouth us?
This is the objection that keeps most of these decisions on hold, and the risk usually runs the other way.
A blunt "we no longer serve accounts your size" email is what damages your reputation, because word travels between customers, and your best accounts start wondering whether they're next.
Options protect you here. Nobody at Sailthru was told to leave. Each customer picked from three choices, and the ones who left made that decision themselves.
It's hard to bad-mouth a company that offered you three ways to stay.
Go deeper
👉 Mostly Metrics: Should You Fire Some of Your Customers? — read CJ Gustafson's full version of the Sailthru story, including how to handle the awkward conversation with customers who are paying you.
👉 MIT Sloan Management Review: Firing the Right Customers Is Good Business — read this before you end any customer relationship; it walks through which customers to keep, fix, or let go, and the trade-offs of each exit route.
👉 SaaStr: How to Fire a Good Client — Jason Lemkin on ending a client relationship without burning a reference you might want for the next decade.
👉 Inc.: Why I Fired My Biggest Client — use this one when the customer in question is a huge share of your revenue; this founder exited a client worth 60% of his and grew 975% in the three years after.
Coming up tomorrow
In tomorrow's issue, you'll learn when to put "AI skills required" within a job role, and what Duolingo's dropped AI rule says about getting AI usage expectations wrong.
That's it for today! Thanks for reading.