Retention, measured rather than assumed
In short
Retention rate follows a fixed cohort; an aggregate repeat purchase rate also moves with new-customer volume. A before-and-after comparison alone does not isolate a programme's effect, nor does a member gap: in our zero-effect example, members still buy 4.3 times as often. Where feasible, use a random control group and count reward costs.
Start with a cohort and a purchase window. The articles explain the metrics, the cost of rewards and how to plan a comparison that can support a business decision.
What loyalty mechanics do to repeat rate, measured rather than assumed.
What this covers
- 01What retention actually measures, and the metrics people confuse for itRead
- 02Raising average order value without buying itRead
Every article in this cluster
Customer retention, and the three metrics people confuse for it
Loyalty campaigns: how to tell if yours did anything
Common questions
Is it really cheaper to retain a customer than acquire one?
Compare the incremental contribution and cost of the available retention and acquisition options over the same horizon. A cost per retained customer needs a credible estimate of how many customers the programme actually retained; it is not a raw count of members.
How do I prove a loyalty programme worked?
Where feasible, randomly assign eligible customers before the offer and compare contribution after programme costs. Size the test for the expected effect and report uncertainty. Other methods need explicit assumptions about selection and concurrent changes.
What is a good repeat purchase rate?
Use the same product category, customer window and metric definition when comparing rates. Your own cohorts help diagnose changes; an external benchmark can provide context only when its population and measurement are comparable.
See it on your data
Loyalz joins ad spend to settled order lines, so the numbers in these articles are columns rather than a spreadsheet you rebuild every month.
Book a demo