Set one decision boundary before comparing scenarios
Records and assumptions to align
- Choose one business, customer cohort or delivery team and one finite horizon. (not complete)
- Use one currency, indirect-tax treatment, time basis and contribution boundary throughout. (not complete)
- Separate observed records, contractual commitments and user-entered scenarios. (not complete)
- State which delivery, support, acquisition and overhead costs are included or excluded. (not complete)
- Hold unrelated inputs constant when testing a sensitivity; do not improve every assumption at once. (not complete)
- Write the evidence trigger and stop condition before selecting a preferred scenario. (not complete)
| Input or boundary | Unit | Evidence class | Required check |
|---|---|---|---|
| Eligible cohort | accounts at start | Billing cohort record | Use one active-state definition |
| Contribution per active account | CU/account/period | Revenue and service-cost record | Use the same horizon |
| Retention intervention cost | CU/cohort | Project or campaign record | Separate fixed and per-account cost |
| Acquisition intervention cost | CU/cohort | Channel cost record | Align attribution and conversion window |
| Incremental retained/acquired accounts | accounts/period | Controlled evidence or scenario | Baseline difference is not causal proof |
Build a reproducible economic view
retention case = incremental retained account-periods × contribution per active account-period - retention intervention cost; acquisition case = incremental acquired account-periods × contribution per active account-period - acquisition intervention cost
- incremental retained account-periods
- Additional active account periods above the selected retention baseline (account-periods/horizon) — comparison evidence or scenario
- incremental acquired account-periods
- Additional active account periods from customers above the acquisition baseline (account-periods/horizon) — attribution evidence or scenario
- contribution per active account-period
- Revenue less the declared cost-to-serve boundary (CU/account-period) — billing and cost record
- intervention cost
- Incremental design, delivery, incentive, media and operating cost for the case (CU/horizon) — project or channel record
Use the same finite horizon, customer-state rule and contribution boundary. Add cash dates and capacity constraints before approving spend.
- Define the decision budget, eligible cohort and finite comparison horizon.
- Reconcile baseline retention and acquisition without the proposed intervention.
- State the incremental account-period scenario for each case.
- Use one contribution-per-account boundary for both cases.
- Include every incremental intervention and service cost.
- Compare lower, base and higher cases plus dated cash timing.
- Choose a measurable test with a stop condition rather than funding an unbounded programme.
Worked example: allocating a 4,000 CU test budget
Invented six-month scenario: 30 CU contribution per active account-month. A retention test costs 4,000 CU and adds 180 account-months; an acquisition test costs 4,000 CU and adds 150 account-months.
| Step | Inputs and arithmetic | Result and interpretation |
|---|---|---|
| Retention contribution before intervention cost | 180 × 30 | 5,400 CU |
| Retention case | 5,400 - 4,000 | 1,400 CU |
| Acquisition contribution before intervention cost | 150 × 30 | 4,500 CU |
| Acquisition case | 4,500 - 4,000 | 500 CU |
| Case | Changed input and arithmetic | Outcome | What to investigate |
|---|---|---|---|
| Retention lower case | 120 × 30 - 4,000 | (400) CU | Check eligible cohort and intervention evidence |
| Base comparison | Retention 1,400 vs acquisition 500 | Retention leads by 900 CU | Still inspect cash and capacity |
| Acquisition higher case | 200 × 30 - 4,000 | 2,000 CU | Validate incrementality and service capacity |
Choose the next test, not a guaranteed answer
| Observed signal | Possible interpretation | Bounded next action |
|---|---|---|
| Churn concentrated in a fixable cohort | A bounded retention test may target the bottleneck | Define cause hypothesis and cohort success measure |
| Retention stable but growth shortfall remains | Incremental acquisition may address the constraint | Align channel CAC and downstream contribution |
| Both cases look positive but cash is tight | Timing may dominate total contribution | Model collection, spend and service-cost dates |
- Comparing saved revenue with acquisition contribution.
- Counting every retained or attributed customer as caused by the intervention.
- Using different horizons or customer definitions.
- Omitting service cost for newly retained or acquired accounts.
- Using an unlimited LTV to justify either case.
Questions to resolve before acting
- Is retention always cheaper than acquisition?
- No. The eligible cohort, intervention cost, incremental effect, contribution, timing and capacity differ by business. Compare measured or explicitly assumed cases.
- What counts as an incrementally retained account?
- An account active beyond the stated baseline under the selected cohort and window. The difference still needs causal caution.
- Should the budget be split between both levers?
- Not by default. A small bounded test for the best-supported bottleneck can create evidence before a larger allocation or split.
Sources and calculation owners
- Subscription Churn Impact methodology — Margin101: Logo and revenue churn scenario owner.
- Customer Acquisition Cost methodology — Margin101: Aligned acquisition-cost denominator and attribution boundary.
- CAC Payback methodology — Margin101: Finite contribution and recovery timing.
Run the next calculation
- Subscription Churn Impact Planner
Keep logo and revenue churn separate while estimating annual contribution and replacement acquisition impact
- Customer Acquisition Cost Planner
Measure loaded and marginal customer acquisition cost across aligned channels
- CAC Payback Planner
Find when retained customer contribution recovers acquisition cost