Loyalty program

TIMi was also used to create churn predictive model for a world-famous reseller of cosmetic products.

The objective here is to detect the customers that won’t buy anymore their cosmetic products at the same re-seller anymore. More precisely: we are actually interested in the opposite prediction: “Who are the recurrent customer?” (in other word, you want a predictive model that predicts who are the customer who will “buy again”). Once you know who your loyal customers are, you can reward them with coupons and price reductions.

Here is the lift of the predictive model:

churnLoyalty

Offering high-valued coupons to non-loyal customers might cost you an enormous amount of money without improving at all the loyalty of your customers.Once again, even in this simplesetting, the accuracy of the predictive model is crucial to the success of your loyalty campaign.