COSTING · CUSTOMER ECONOMICS
Your largest customer may not be your most profitable
Revenue rankings can hide delivery effort, small orders, returns and support time. Build a decision view of customer contribution without pretending every allocated cost disappears if the customer leaves.
THE SHORT ANSWER
Begin with net sales and directly attributable product or service costs. Add evidenced customer-service activities using clear drivers, label shared-cost allocations and test which costs would actually change under the proposed action.
Choose the decision before building the allocation model
A customer report should help answer a specific question: change a delivery pattern, quote a new contract, review an order minimum or invest in account support. It should not simply replace a sales ranking with a profit ranking. Use the same period, currency and customer definition across invoices, credits, deliveries and support records. Exclude sales taxes collected for authorities where appropriate to the management measure and reconcile the totals to the finance records.
Start with net sales after evidenced discounts and credits. Identify product or service costs already included in gross margin. Then trace service activities that vary meaningfully between customers. Keep an explicit boundary between cost already counted and cost added in this analysis. Counting delivery labour in product cost and again in a service pool can turn a useful comparison into a misleading loss.
ACCA’s activity-based management material explains using activity drivers to examine customer economics. BDC’s customer-profitability guidance provides a separate business perspective. The decision record and numbers below are original Mizwala illustrations, not figures taken from either organisation.
Compare two customers on a consistent basis
In a fictional quarterly review, Customer A generates OMR 60,000 net sales and OMR 18,000 gross contribution. Customer B generates only OMR 40,000 sales and OMR 12,000 gross contribution. Both begin at 30%. A revenue-only review makes A look much more valuable.
The service records change the discussion. A has 100 deliveries at an estimated OMR 45 per delivery, 120 support hours at OMR 35 per hour and OMR 1,800 attributable returns-handling cost. B has 30 deliveries, 20 support hours and OMR 250 returns-handling cost using the same definitions. Contribution after these activities is OMR 7,500 for A and OMR 9,700 for B. These are management measures before unallocated shared overhead, financing and tax; they are not statutory net profits.
The table is not evidence that A should be dropped. It is evidence, within the invented scenario, that delivery frequency and support demand deserve investigation. Validate the rates, sample size, capacity assumptions and double-counting risk before a real decision.
| Measure | Customer A / Customer B |
|---|---|
| Net sales | 60,000 / 40,000 |
| Direct product/service cost | 42,000 / 28,000 |
| Gross contribution | 18,000 / 12,000 |
| Delivery activity | 4,500 / 1,350 |
| Support activity | 4,200 / 700 |
| Returns handling | 1,800 / 250 |
| Contribution after listed activities | 7,500 / 9,700 |
A: 18,000 − 4,500 − 4,200 − 1,800 = 7,500. B: 12,000 − 1,350 − 700 − 250 = 9,700. Fictional estimates; before remaining shared costs, financing and tax.
Separate measured effort from avoidable cash cost
Use drivers you can explain: deliveries, order lines, service hours or return cases. A practical first model might use a measured sample of support time rather than a new tracking system. Record how the sample was chosen, missing records and the uncertainty in each rate. A blended cost per hour is an estimate; displaying two decimal places does not make it precise.
Keep customer contribution, allocated shared overhead and incremental decision economics in separate columns. If a salaried team remains after a customer leaves, its allocated cost is not immediately saved. If freed capacity can serve another profitable account, that opportunity may matter even when payroll cash does not change. Any forecast of replacement demand needs its own evidence.
Look at payment timing and concentration separately. A profitable customer can still consume cash through late payment; a low-margin customer may share delivery capacity with other accounts. For a proposed change, inspect contractual commitments, service quality, relationship value and what happens to the rest of the route or team. Avoid scoring customers using assumptions you would not be willing to explain internally.
Test a service change before issuing a new price
For fictional Customer A, reducing deliveries from 100 to 60 at the same estimated OMR 45 rate reduces the attributed delivery measure by OMR 1,800. The modelled contribution rises from OMR 7,500 to OMR 9,300 if everything else is unchanged. Actual cash savings could be smaller if routes, staffing or vehicle commitments remain. Confirm that fewer deliveries would still meet the customer’s needs and agreed terms.
Other options include consolidating orders, reducing preventable returns, improving ordering instructions or pricing an additional service explicitly. Use the pricing calculator to explore a price change and the receivables guide to examine collection timing. Review any revised offer through the normal commercial approvals.
Agree a limited trial, an owner and a review date. Compare actual service effort, customer response, contribution and cash with the baseline. Keep a relationship decision broader than a single-quarter allocation report. The worksheet below records the evidence needed to challenge the model before acting on it.
TAKE THE NEXT STEP
Use the worksheet.
Open the worksheet, type your notes in the fields and use your browser’s Print command to keep a PDF or paper copy. Calculations are manual. Entries stay in your browser and are not submitted to Mizwala; reload or close without saving and they may be lost.
Questions worth checking
Is contribution after service the same as net profit?
No. State which costs are included and excluded. Shared overhead, financing and tax may remain outside the measure.
Should we stop serving the lowest-ranked customer?
Not from the ranking alone. Validate the model and consider avoidable costs, capacity, contracts, payment timing and the wider relationship.
Sources & scope
Primary references checked on 4 November 2026. The calculations, examples and workflow suggestions are Mizwala’s educational illustrations; the linked organisations do not endorse this guide.
This is general management information. Adapt it to your circumstances and obtain appropriate professional advice before relying on it for a material decision.