The short answer
Customer profitability is what a customer earns you after the cost of serving them, not just after the cost of goods. To measure it, start with gross margin by customer from your invoices, then subtract the costs that vary by customer: order handling, delivery, returns, credit terms, rebates, and sales time. When companies do this, the usual result is a whale curve: Kaplan and Narayanan found the most profitable 20 percent of customers can generate 150 to 300 percent of total profit, while the least profitable customers erase much of it17.
Why gross margin misleads
Two customers can buy the same products at the same margin and be worth completely different amounts. One orders twice a month in full pallets and pays in thirty days. The other orders every other day in single cases, wants rush delivery, returns a tenth of what it buys, and pays in seventy. Gross margin says they are identical. Your warehouse, drivers, and controller know they are not.
Step 1: gross margin by customer
Pull twelve to twenty-four months of invoice lines with customer, product, quantity, price, and cost at the time of sale. Add rebates and credits back against the customer who earned them. This is the starting point, and it is often the first time the business has seen it clearly.
Step 2: add the cost to serve
Cost to serve is every cost that rises or falls with how a customer buys. You do not need an activity-based costing project to estimate it. A few drivers explain most of the difference between customers.
- Orders and lines: each order costs roughly the same to pick, pack, and invoice, whatever it is worth.
- Deliveries: stops, distance, and rush or special shipments.
- Returns and credits: handling cost plus the margin given back.
- Payment terms: the cost of carrying the receivable for the days the customer takes to pay.
- Sales and service time: rep visits, quotes that never convert, and support calls.
Estimate a cost per unit of each driver from your P&L, such as the cost of one order line or one delivery stop, then multiply by each customer’s activity. The estimate does not need to be exact. It needs to be applied the same way to every customer.
Step 3: rank and read the curve
Sort customers from most to least profitable and plot cumulative profit. The curve usually climbs well above 100 percent and then falls back as the unprofitable customers are added. That shape is the finding. The top of the curve shows who to protect. The tail shows where the business is paying to keep revenue.
Specimen · illustrative figures
A $35M industrial supplier had 1,400 active customers. After cost to serve, the top 280 produced 190 percent of operating profit. The bottom 350 lost a combined $620K, mostly through small frequent orders, free delivery, and 60-day payment. Changing minimum order rules and freight terms for that group was worth an estimated $300K to $450K a year.
What to do with the tail
Firing customers is rarely the answer. Most unprofitable customers can be made profitable by changing how they buy rather than whether they buy: a minimum order, a delivery schedule, a charge for rush orders, or a move to online ordering. Price comes last. Test each change on part of the tail against a control, agree in advance what result ends the test, and watch whether the customers stay.
Write the hunch down first
Before anyone runs the numbers, ask the leadership team to name the five customers they think are most and least profitable. The list they write and the list the records produce rarely match, and the differences are where the conversation should start9.
Sources
- 17Kaplan, R. S. & Narayanan, V. G. (2001). Measuring and managing customer profitability. Journal of Cost Management, 15(5), 5–15.
- 9Fischhoff, B. (1975). Hindsight ≠ foresight. Journal of Experimental Psychology: Human Perception and Performance, 1(3), 288–299.
Market figures describe the landscape and are not Modven results.