The DMA-CPA approach is a proposed budgeting heuristic: estimate the contribution available from a customer, reserve part of it, and use the remainder as an acquisition-spend ceiling. I find that more useful than chasing the lowest possible CPA without asking what a customer is worth. It is not a validated forecasting model or a guarantee of profitable growth.
Substantively reviewed and updated: 22 October 2026. The examples below are hypothetical and illustrate the arithmetic, not measured business results.
Start with a defined cost and time horizon
Observed CPA is acquisition spend divided by the number of qualifying acquisitions. Define the acquisition first: a lead, order and new paying customer are different outcomes. In this article, the denominator is new paying customers and the numerator includes all the acquisition costs being assessed. A media-only CPA is narrower than a fully loaded customer acquisition cost.
For an allowable acquisition budget, use expected customer revenue over a stated horizon and subtract the costs of serving that revenue. I use “pre-acquisition contribution” to mean revenue after the specified delivery, service and retention costs, but before the new acquisition spend being evaluated. List those costs explicitly: production, fulfilment, payment fees, support, refunds and ongoing retention marketing may all matter. Do not use a margin that already deducts the same acquisition spend and then deduct it again.
Gross profit and gross profit margin also differ. Gross profit is revenue minus cost of goods sold; gross profit margin is that amount divided by revenue. Neither automatically captures every cost relevant to an acquisition decision.
The calculation
Let R be expected revenue per acquired customer over the chosen horizon, m the pre-acquisition contribution margin as a fraction of that revenue, and b the contribution reserve as a fraction of the same revenue. Then:
Allowable acquisition spend per customer = R × (m − b)
For example, m = 0.50 and b = 0.20 leave 0.30 of revenue for acquisition. The reserve is 20 percentage points of revenue, not 20% of the 50% margin. I previously called the remainder Maximum Allocatable Profit; “acquisition allowance” makes its purpose clearer.
This shortcut assumes the margin describes the same customer cohort and horizon as the revenue. If expected costs vary substantially over time, calculate the expected contribution directly instead of multiplying revenue by one margin. If your LTV already represents expected discounted pre-acquisition contribution, use that amount directly and subtract a consistently discounted reserve; do not multiply it by another margin. If m is no greater than b, this rule supplies no positive acquisition allowance; revisit the economics rather than treating a negative result as an executable budget.
Hypothetical example: bottled water
HydraQuench expects a customer to buy 30 packs at $6 each over one year: R = $180. This is one-year revenue, not an estimate of lifetime value. Assume its defined pre-acquisition contribution margin is 50% and it reserves 20% of revenue.
$180 × (0.50 − 0.20) = $54 available for acquisition. Under those assumptions, $90 of pre-acquisition contribution minus $54 of acquisition spend leaves a $36 reserve. Longer retention could change the calculation, but must be estimated rather than inferred from annual spending.
Hypothetical example: SaaS
CloudConnect assumes a customer pays $20 monthly for 36 months: R = $720. For this simplified, undiscounted example, assume a 60% pre-acquisition contribution margin and a reserve of 20% of revenue. These are illustrative inputs, not typical SaaS benchmarks.
$720 × (0.60 − 0.20) = $288 available for acquisition, leaving $144 from $432 of expected contribution. In practice, model cohort retention and the timing of costs and receipts: an average three-year subscription assumption hides churn and cash-flow risk. A lifetime allowance can be unaffordable if acquisition must be paid today and recovered slowly.
Hypothetical example: consulting
Strategize Group assumes a client pays $50,000 annually for two years: R = $100,000. Its assumed pre-acquisition contribution margin, after the defined delivery costs, is 15%; the reserve is 10% of revenue.
$100,000 × (0.15 − 0.10) = $5,000 available for acquisition. That leaves a $10,000 reserve from $15,000 of contribution. Include proposal preparation, sales time and commissions in the acquisition cost definition where relevant. Scope changes, unpaid invoices and staff capacity can alter the result.
Turning a ceiling into a decision
- Build estimates from comparable customer cohorts, with a documented revenue horizon and cost schedule. Use conservative scenarios for retention, refunds and service costs.
- Decide whether the reserve must also cover fixed overhead, financing, tax and profit. A contribution reserve is not automatically net profit. Apply a separate cash-payback constraint when needed.
- If the ceiling covers all acquisition costs, subtract non-media acquisition costs before setting a media-only target. Keep the same denominator when comparing target and observed cost.
- Compare marginal performance as spend increases. A profitable historical average does not establish that the next increment of spend will perform equally well.
- Review assumptions when prices, costs or customer mix change. Report uncertainty and test outcomes; do not raise the ceiling simply to make an expensive campaign appear viable.
The heuristic sets an affordability constraint under stated assumptions. It does not identify which channel caused a purchase or which budget will maximise profit. For those questions, see experimental approaches to social media ROI and what marketing mix modelling can measure. Google’s Meridian documentation distinguishes historical channel contribution, response curves and budget optimisation—questions this arithmetic alone cannot answer.