The company managing different portfolios of private label cards needs tools to identify, on a recurring basis, the customer propensity for returning to each store. Thus, it is possible to act in a preventive manner by observing signals sent by the consumers, which reduces the chances of losing the client due to disinterest in the card.
This process comprised the analysis of new and old customers in retail and wholesale. The models were implemented for continuous use and make it possible to act on each portfolio and each client, who is linked to his or her own probability of remaining active in the card base.
