Your largest customer may use the most support: measure cost to serve

Two customers can buy the same amount at the same product margin and leave very different amounts to cover the rest of the business. One sends a few complete orders. The other needs frequent order changes, delivery investigations and invoice corrections. A revenue-based overhead allocation treats them alike precisely where their behavior differs.

A cost-to-serve review assigns selected service costs using the work customers consume. Its immediate purpose is to identify a service pattern management can change, not to declare that an allocated customer loss would disappear if the account left. ACCA describes customer profitability as one application of activity-based management.

Choose the decision before collecting more data. Perhaps the account team wants to consolidate small orders, change the support entitlement or investigate why one customer creates repeated corrections. Select a completed period long enough to contain that work and keep the same customer population throughout the calculation.

Begin with revenue and costs already traced to customers: net sales, the agreed product-cost measure, freight and other direct charges. Then identify service costs missing from that view. Mark whether each expense is already included in product cost or another allocation. The same support salary must not enter both the product margin and the new service pool.

Keep unrelated company costs separate. A review of order administration should not acquire the chief executive's entire salary simply because it has nowhere else to go. Show the excluded balance explicitly rather than promising that this limited study calculates each customer's complete economic value.

Group service costs by work that can be observed, then choose a relevant measure for each group. Order processing might use completed orders; returns administration might use return cases. When case complexity varies substantially, handling time or separate case categories may be more useful than an undifferentiated count.

This follows the activity-pool and driver-rate structure explained in OpenStax's managerial accounting text. The following worksheet is a hypothetical customer-service application, not a published benchmark.

Each rate divides its own cost pool by its own activity count. Do not combine 600 orders, 600 contacts and 180 returns into one denominator: they represent different work.

These are average assigned rates for the selected period. They do not establish what the next order costs in cash, and a support contact does not become a $20 invoice merely because the analysis assigns it $20.

Suppose Customers A and B each generated $100,000 of net sales in the quarter. After the same $60,000 product cost and $3,000 directly traced delivery cost, each shows $37,000 before the service pools above.

Their recorded service activity differs:

For A, the calculation is 30 × $15 + 10 × $20 + 2 × $50. B's calculation uses the same rates and its own quantities. The remaining amount after these assigned service costs is $36,250 for A and $32,200 for B, a $4,050 difference that the initial product-and-delivery view concealed.

Neither result is company net profit. Other costs remain outside this deliberately limited comparison. The useful finding is specific: B creates more work in all three measured activities. The account owner can now investigate the causes rather than arguing about a blanket overhead percentage.

Complete the reconciliation across all customers. In this example, A and B receive $5,550 of the $30,000 pool. The remaining customers must account for the other $24,450 under the stated allocation method, or the worksheet must identify an unassigned balance. Preserve the source counts with the calculation.

A customer identifier on a support ticket is not enough to establish its cause. Separate ordinary service from corrections caused by your own billing or delivery errors. Charging a customer for avoidable internal mistakes may be the wrong commercial response even when the work is correctly measured.

Check duplicates, reopened cases, shared accounts and incomplete records. If one customer receives phone support that never reaches the ticket system, its apparent advantage may reflect missing data. When records are sampled or time estimates are used, label them and test whether plausible changes would alter the account comparison.

Do not make a complex case and a two-minute query equivalent merely to keep the spreadsheet simple. Conversely, avoid building dozens of categories that staff cannot maintain. Add detail where it could change the proposed action.

Suppose B agrees to batch orders and use a clearer returns process. Fewer tasks may free staff capacity without reducing salaries or other committed payments. The revised activity allocation can show less cost assigned to B even while company spending stays unchanged.

A cash-saving claim therefore needs a further question: which future payment changes, by how much and when? ACCA's relevant-cost guidance distinguishes decision-changing cash flows and opportunity costs from costs that continue regardless. Apply that distinction before treating a service redesign or customer exit as a saving.

Choose a limited action with the commercial owner. It might be an order-consolidation trial, corrected invoice instructions or a separately priced service level, subject to the actual agreement. Record the expected change in activity, the cost that could genuinely move and any service-quality measure that must be protected.

At the review date, compare those observations with the original worksheet. A successful change should be visible in the work and the customer relationship, not only in a different allocation formula.