First-order economics: returns, acquisition cost and cash

First-order contribution after acquisition equals a new customer's initial net order revenue less the costs of that order and the acquisition spending attributed to that customer. Use a matched acquisition cohort. Keep later purchases separate so an assumed future order cannot hide a loss on the first one.

Build a matched customer and order group

Assume a campaign acquires 100 new customers and each places one initial delivered order. The first-order denominator is therefore 100. Ad spending is $1,800. Do not divide it by all store orders: if 50 orders also came from existing customers, $1,800 ÷ 150 = $12 is not this cohort's acquisition cost. $1,800 ÷ 100 = $18 per acquired customer.

Specify which ad and acquisition costs are included, how a customer is assigned to the cohort and how long returns are observed. An attribution rule allocates credit; the resulting ratio alone does not demonstrate that every customer was caused by the spending. Compare like-for-like cohorts before shifting a budget.

Include the separate effects of returns

For the example cohort, assume customer charges before discounts are $7,000, discounts are $200 and refunds are $1,000. Product cost shipped is $2,600, of which $400 returns to saleable stock at the same cost. Return handling costs another $300. Amounts exclude sales tax collected for others.

The refund reduces revenue. Recoverable inventory reduces the net goods cost. Handling is an extra expense. Keep all three effects, but count each once. Do not subtract the $1,000 refunds again below net revenue, or treat all returned goods as worthless when the $400 stock recovery is actually supported.

Calculate the acquisition ceiling

The cohort leaves $23 per customer before acquisition. A $23 acquisition cost consumes all of that contribution, leaving nothing for shared overhead. At $24 it loses $1 per customer, or $100 for this cohort. If the first order must contribute $7 toward overhead and profit, the acquisition ceiling is $23 − $7 = $16, under the same order assumptions.

This is an operating screen, not an instruction to bid exactly $16 in an ad platform. A different channel can change discounts, returns, basket size and delivery cost as well as acquisition cost. Recalculate the full bridge for the customers that extra spending is expected to bring.

Reconcile contribution with cash

Assume the cohort's payments all occur within the analysis window. Customers initially pay $6,800 after discounts, then receive $1,000 refunds. The business pays $2,600 for the shipped stock, $700 fulfillment, $300 fees, $300 handling and $1,800 acquisition. Net cash movement is $100.

Contribution was $500. The $400 difference is returned stock still held at cost, not cash collected from selling it. Crediting the inventory recovery as an immediate bank receipt would overstate liquidity. Put the actual ad, stock, customer-payment and refund dates into the cash forecast; even the $100 net figure can hide an earlier funding gap.

Make repeat-purchase recovery an explicit second calculation

At $24 acquisition cost, the first order loses $1. Suppose, only as a scenario, that 25% of the cohort places a later order contributing $8 before any new acquisition spending. Expected repeat contribution is 25% × $8 = $2 per original customer, leaving $1 after the first-order loss. The timing of that repeat receipt still matters.

That result depends entirely on the repeat rate, order contribution and observation period. Use mature customer evidence rather than inserting one guaranteed repeat order for every customer. The cohort forecast guide shows why retention by age matters; a one-time retail repeat purchase has its own behaviour and should not inherit a subscription churn assumption.

For the cost definition behind contribution, see the delivered-order margin guide and SBA's contribution-based break-even explanation. Use the weekly cash schedule for the cash needed before acquisition pays back. The values above are constructed teaching inputs, not return-rate or acquisition-cost benchmarks.