Order Fulfillment Cost Calculator

The calculation itself

Order fulfilment cost in its complete form is everything between the order being placed and the money being finally kept: pick and pack, materials, shipping, payment processing, the share of orders that generate a support contact, the share that return, and the share that charge back.

It is the cost-to-serve figure rather than the warehouse figure, and it is the one that determines whether an order was actually worth taking.

Running the numbers

The standard $58 order. Goods $26.10. Pick and pack $2.40, materials $0.95, shipping $9.85, $13.20 of physical fulfilment.

Payment processing at 2.9% + $0.30 is $1.98. Support: 18% of orders generate a contact averaging 6 minutes at $24 an hour, adding $0.43. Returns at 12% costing $8.60 each adds $1.03. Chargebacks at 0.3% costing $73 including the fee adds $0.22.

Total cost to serve is $16.86, against $13.20 of physical fulfilment, 28% higher. Contribution after everything is $58 − $26.10 − $16.86 = $15.04, or 26% of revenue, before any advertising or overhead.

What gets missed

The physical fulfilment number gets used as the fulfilment cost because it is the one the warehouse reports. Payment fees sit in finance, support sits in customer service, and returns sit somewhere else again, so no single report contains the true figure.

Averages also hide the tail badly. A small share of orders generate multiple support contacts, a return, and a partial refund, and those orders can cost several times the average while looking identical in the aggregate.

What to do next

Build the full figure once, from four different sources if necessary, and then use it as the basis for every acquisition decision. An advertising budget set against gross margin rather than cost-to-serve overspends by whatever the difference is.

Then rebuild it per channel. Marketplace orders, direct orders and wholesale orders have materially different cost-to-serve, and a single blended figure will make one of them look better than it is.

Cost to serve by customer, not just by order

The same arithmetic applied per customer over a year is more revealing still, because the costs are not evenly distributed. A customer who orders four times, returns once and contacts support twice has a very different cost profile from one who orders four times and never gets in touch.

Ranking customers by contribution after cost-to-serve typically shows a familiar shape: a profitable majority, a thin tail of highly profitable repeat buyers, and a small group who are reliably loss-making, usually through returns rather than through anything else.

That last group is worth identifying because the response is a policy rather than a punishment: a restocking fee above a return threshold, better sizing information, or simply not remarketing to them. Businesses that never build the view end up spending acquisition budget attracting more of exactly the customers they lose money on.

One practical note on building the figure: the four sources rarely reconcile to the same order count, because payment records, shipping records, support tickets and returns each use a different identifier. Joining them on the order number is worth the effort once and impossible to do casually each month.

Building it as a scheduled report rather than a one-off analysis is what turns it from an interesting exercise into a number the business actually manages against.