Choose the recovery measure
For a subscription business, contribution may be observed over recurring periods after relevant variable costs. For commerce, it may accumulate across purchases. Define the included costs and whether refunds, failed delivery or payment losses affect the measure.
Historical payback and modeled payback are different. A model uses assumptions about future activity; observed recovery uses actual customer outcomes. Label them separately.
Build the estimate
- Define acquisition cost per eligible customer or cohort.
- Calculate contribution over comparable periods.
- Include the delay before revenue is collected or purchases recur.
- State retention and margin assumptions for future estimates.
- Monitor actual cohort recovery against the model.
A simple division can be useful when contribution is reasonably stable, but it should not conceal changing customer behavior.
Worked example: a simplified model
Illustrative example: acquisition cost is 600 units, and expected monthly contribution is 50 units under a steady-use assumption. Simplified modeled payback is 600 / 50 = 12 months. If customers leave earlier or contribution changes, actual recovery differs.
The calculation is not a promise that every customer repays acquisition spending. Show the underlying retention and cost assumptions, especially for new cohorts with little history.
Payback decision table
| Issue | What to inspect |
|---|---|
| Slow recovery | Acquisition cost, contribution and time to first payment. |
| Strong average, weak segment | Customer mix and segment-specific economics. |
| Modeled recovery, little history | Sensitivity to retention and margin assumptions. |
| Revenue recognized before cash collected | Cash timing and operational requirements. |
Connect payback to cash capacity
A business may have attractive long-term economics but insufficient cash to fund a long recovery period. Marketing allocation needs to respect that constraint. Treat the model as a planning input and involve the appropriate finance owner for consequential decisions.
Is a shorter payback always preferable?
It reduces recovery delay, but the business also needs customer value and capacity context. Avoid judging different customer segments from one number alone.
Should lifetime value replace payback?
No. Lifetime estimates and recovery timing answer different questions. Early lifetime estimates can be especially uncertain.
Put this into practice
Create an observed contribution curve for one mature cohort and a clearly labeled model for a newer one. List the assumptions that would change the recovery conclusion most strongly.
Related foundation: CAC vs ROAS: which metric should guide your marketing budget?. How these guides are prepared.
Ayoub Mouhachtt
Growth & performance marketing. Explore the portfolio and working background.
Related portfolio work: Etsy & Shopify. The worked examples in this guide are illustrative and are separate from the portfolio’s project evidence.