The challenge
Customer acquisition cost had been optimised, the brand was good at getting customers. What happened after the first purchase was not managed at all. Repeat purchase rate was 22%, which is below-average for their category. The team knew this but had no system to address it.
What we built
We built a customer health model based on purchase recency, frequency, and category overlap. Customers in the "at risk" tier received a personalised WhatsApp sequence at the point of predicted churn, before they were already gone.
Repurchase triggers were built for the highest-frequency category combinations. Win-back sequences ran automatically for customers inactive for 90 days.
The outcome
Repeat purchase rate moved from 22% to 34% in the first six months. Win-back sequences converted 18% of lapsed customers. Customer LTV increased 28% across the active base, without any change to acquisition strategy or product.