Repurchase Prediction model

What is the Repurchase Prediction model?

While the Propensity to Buy model gives a general 30-day purchase probability, the Repurchase Prediction model specifically identifies customers who are statistically overdue based on their personal purchase rhythm.

A customer with a 0.82 repurchase score and who hasn't bought in 45 days (when their average gap is 38 days) is a strong candidate for a gentle nudge — they intend to buy again, they just need a reminder.

Repurchase score vs. win-back window

• High repurchase score → Customer is likely to buy again — they just need a nudge at the right time
• Win-back window open → The right time is now — receptivity is at its peak
• Both → Highest-priority target: a customer who wants to buy and is currently receptive