Case study 07 / Retention strategy / Unit economics
Cure.fit Retention & Revenue Quality
Strategic model
Retention is an economic problem, not just a marketing metric.
An independent strategic model connecting retention, utilisation, customer lifetime value and acquisition economics.
Editorial illustration of the project’s focus.
01 / Context
The problem
Acquisition spending creates fragile growth when customers leave before the business recovers the cost of winning them.
02 / Ownership
My contribution
Financial modelling, churn segmentation, retention analysis, channel strategy and CAC sensitivity analysis.
The modelling approach
- Connect retention and utilisation with break-even economics.
- Segment churn to identify different intervention needs.
- Test acquisition-cost sensitivity against utilisation assumptions.
What the data can establish
The workbook uses an illustrative proxy sample and explicit churn assumptions. It separates observed sample behaviour from modelled monthly churn and warns against treating an annual cohort’s zero observed churn as a population finding.
Cohort differences are not the same as achievable intervention upside. Retention scenarios need to account for selection effects, the share of members who could change behaviour and the assumptions behind the LTV calculation.
A proposed acquisition channel
The strategy considers B2B payroll-embedded acquisition as a route to customers. The proposal is evaluated as a commercial scenario rather than an implemented company programme.
Revenue quality
The decision is not simply how cheaply a customer can be acquired. It is whether the customer uses and retains the service long enough for the relationship to create economic value.
The idea to carry forward
Acquisition becomes more valuable when the business can retain customers long enough to recover the cost of winning them.
Independent strategic modelling using illustrative proxy data. Model outputs, proposed channels and retention scenarios are not actual Cure.fit company results or verified member-level findings.