Case Study
Customer Analytics
Customer Lifetime Value
How can the future value of each customer relationship be estimated to inform decisions?
Download PDF
A complementary visual resource for a deeper view of the case.
The business challenge
What economic value might a future relationship generate, with what uncertainty, and for which decisions is it sufficiently reliable?
CLV connects acquisition, retention and service with expected value. A number without a horizon, costs or uncertainty can lead to poor priorities.
Key business questions
- Which future value should be estimated, and over what horizon?
- Which revenue, margin, cost and tenure components matter?
- Which decisions will use the estimate?
Data & evidence
- Purchases, revenue, margin and returns.
- Frequency, recency and tenure.
- Service, acquisition and incentive costs.
- Product, channel and cohorts.
How I would approach it
- Define horizon, economic unit and use.
- Separate frequency, value and tenure.
- Validate on future cohorts and periods.
- Represent uncertainty and sensitivity.
What you would get
- CLV estimate with explicit horizon and assumptions.
- Customer value distribution and profiles.
- Drivers, uncertainty and scenarios.
- Recommendations for acquisition, service and retention use.
From analysis to action
Illustrative example: CLV is an expectation conditional on data and assumptions, not a revenue promise.
- Adjust acquisition and service thresholds.
- Prioritise strategies by value and need.
- Inform segmentation and planning.
How success should be measured
- Error and calibration by horizon.
- Ranking stability.
- Incremental value of decisions informed by CLV.
Other approaches to consider
Historical, BG/NBD, survival or supervised models may be appropriate. Complexity should be justified by a better decision.