Customer retention and customer acquisition: compare the economics

Customer retention and customer acquisition should be compared using customer contribution, consistent costs and the same observation horizon. Acquisition wins new relationships; retention supports relationships already in the customer base. Neither deserves more budget simply because its cost per customer looks lower.

What each investment measures

A US commercial supplier needs both a flow of new accounts and profitable repeat business. Its marketing, sales and account-management budgets overlap, so finance must assign spending to a stated purpose before comparing returns.

For the acquisition denominator and cost definition, use customer acquisition cost. For the other budget, cost of customer retention owns the detailed cost calculation. Here the question is how to put both investments on a comparable economic basis.

Bring costs and customer returns together

Use the customer master, CRM outcomes, invoice export, credits, direct service or product costs, and spending records. Customer IDs should connect each invoice and cost to the appropriate customer group. Record a separate allocation for shared labor rather than charging it to both budgets.

Use net revenue after returns and discounts. In this comparison, variable fulfillment costs include the product or service costs that move with activity. Acquisition and retention program spending is separate, so it is deducted once.

Customer contribution before program spend = Net revenue − Variable fulfillment costs

Observed contribution after program spend = Customer contribution before program spend − Assigned acquisition or retention spend

The SBA's break-even guidance distinguishes selling price, variable costs and contribution. For this internal decision, finance must document exactly which costs it treats as variable and which program expenses sit outside them. The resulting contribution is not company net income: unallocated corporate overhead, taxes and financing costs remain outside the example.

A worked comparison on the same rows

The supplier compares a new-account cohort that began purchasing on January 1, 2026, with an existing-account group present on that date. Both groups contain 100 customers and are observed through March 31, 2026. Starting both windows together avoids giving one group more time to produce revenue.

The existing group ends with 85 active customers. Contribution includes all eligible activity from its original 100 customers during the period, including purchases from accounts that later became inactive.

For acquisition, 240,000 minus 144,000 = $96,000 of contribution before program spending, and 96,000 minus 60,000 = $36,000 afterward. For the existing base, 320,000 minus 192,000 minus 40,000 = $88,000 afterward.

The reconciliation is 560,000 minus 336,000 minus 100,000 = $124,000, which also equals 36,000 plus 88,000. Revenue, fulfillment cost and program spending must each tie to the assigned source rows and the corresponding finance totals.

Acquisition spending per new customer is 60,000 / 100 = $600. Retention spending per eligible existing customer is 40,000 / 100 = $400. Dividing by the 85 customers active at the end instead gives $470.59 per observed retained customer, rounded to cents. Neither retention denominator measures customers saved by the spending.

Customer retention is 85 / 100 = 85.0%, shown to one decimal place. The cohort retention table guide explains how to maintain the fixed customer membership behind such a comparison.

Do not turn observed returns into incremental ROI

The existing base produced more contribution after its assigned program costs. That is useful descriptive evidence. It does not mean canceling retention spending would remove all of that contribution, or adding retention spending would reproduce the same return.

Established customers may buy without intervention. New customers may produce further contribution after the observed window. A longer comparison needs actual later purchases or explicitly labeled assumptions; the customer lifetime value guide covers the historical customer-return side.

For the next budget decision, compare incremental proposals. The following quantities and contribution per customer are planning assumptions for the same future observation period, not measured program effects.

The retention proposal assumes 12 genuinely additional customers preserved: 12 times 1,200 minus 10,000 = $4,400. Acquisition produces 10 times 1,400 minus 10,000 = $4,000. The expected advantage is only $400; halving the retention effect changes its result to a $2,800 loss.

The commercial owner must justify those additional-customer assumptions. A suitable comparison group or randomized test can strengthen the estimate; NIST's experimental-design guidance explains random assignment. Accounts targeted because they already looked likely to leave cannot simply be compared with an easier untreated group.

Where the comparison goes wrong

• Different horizons. Mature existing accounts are compared with new accounts that have barely started purchasing.
• Revenue used as return. High sales with expensive fulfillment can contribute less than a smaller, lower-cost account base.
• Costs deducted twice. Discounts already reduce net revenue; shared service expenses cannot also be charged in full to both programs.
• Retention credited for all survivors. Observed retained customers include those who would have stayed without the program.
• Average economics used for the next dollar. A profitable historical cohort does not establish the return from an additional campaign or service commitment.

Compare your customer economics in Covirage

Bring your customer, invoice, fulfillment-cost and commercial-spending exports; Covirage maps the columns, and its tools compute contribution comparisons and check the totals. The external AI model explains the results with every figure traced to its rows. Create a free Covirage account and bring a sample of your own data.

See customer profitability for the supporting account-economics use case.