Do we need activity-based costing software for this?

No. Four cost drivers priced from your own P&L are enough to rank customers and size the decision. Precision matters less than a consistent method.

Will a minimum order fee drive customers away?

Some may consolidate orders, which is the point. The test sizes the change by segment first, and the rollout starts with one branch against a holdout.

Small-order cost to serve in industrial MRO distribution: how to find it in your records

The short answer

MRO distribution runs on many small, urgent orders, and each one carries picking, delivery, invoicing, and often a credit or return. Add those costs to every order and many small orders lose money even at good gross margin. Rank customers by profit after cost to serve and plot the whale curve. If a group of customers destroys part of the profit the rest create, minimum order rules, delivery fees, or a vending program usually win it back.

Where it hides in industrial MRO distribution

Emergency and will-call orders

A single bearing delivered same day can cost more to deliver than it earns.

Orders split across days

A customer who orders the same items in five small orders a week costs five times as much to serve as one who orders once.

Returns and credits

Wrong part, no longer needed, or bought elsewhere. Each return doubles the handling and erases the margin.

Payment behavior

Small accounts on net 30 that pay at 75 days carry a financing cost nobody charges for.

The records to pull

Twelve to twenty-four months of each. Nothing needs to be installed or connected; a spreadsheet export is enough.

Invoice lines

Customer, ship-to, order date, ship method, line count, revenue, and cost.

Delivery and route data

Stops, routes, carrier charges, or fleet costs, even if only at route level.

Returns, credits, and AR aging

Credit memos by customer and days to pay by invoice.

The ten-day test

Days 1–2

Ask the team which customers it believes are least profitable, and write the list down. Hold out the last quarter.

Days 3–5

Build a cost-to-serve model with four drivers: orders, lines, deliveries, and returns, priced from your own operating costs.

Days 6–8

Rank customers by profit after cost to serve and draw the whale curve. Compare the result with the team’s list.

Days 9–10

Confirm the ranking in the holdout and size three changes: minimum order value, delivery fee below a threshold, and vending for high-frequency accounts.

Specimen · illustrative figures

A $110M MRO distributor modeled cost to serve on 2,300 customers. The most profitable 30 percent of customers produced 165 percent of operating profit, and the bottom 25 percent destroyed 48 percent of it. Orders under $150 were 44 percent of order count and 7 percent of revenue. Only four of the team’s ten suspected money-losers were actually in the bottom tier.

Stop condition, written first

Stop if no customer group loses money after cost to serve in both the search period and the holdout, or if the losses are too small to cover the cost of changing terms.

What gets built if it holds

Customer-level profitability refreshed monthly, with order minimums and delivery fees applied by segment, and vending or scheduled replenishment offered to high-frequency accounts. It runs inside your operation, belongs to you, and is measured weekly against the dollar range the test produced.

Sources

17Kaplan, R. S. & Narayanan, V. G. (2001). Measuring and managing customer profitability. Journal of Cost Management, 15(5), 5–15.

Specimen figures are illustrative and do not describe a Modven client. Reviewed October 10, 2026.

Small-order cost to serve in industrial MRO distribution

Where it hides in industrial MRO distribution

The records to pull

The ten-day test

What gets built if it holds

Asked often, answered plainly.

Where do you suspect the value is hiding? Tell us your hunch. Ten days later it comes back with numbers on it.