karagul / bank-loss-warning-model Goto Github PK

This project forked from heiheiwangergou/bank-loss-warning-model

Bank churn refers to the bank's customers terminating all business in the bank and selling them. However, in actual operations, for a specific business unit, bank customer churn can be positioned as a specific business termination behavior. The loss of customers in commercial banks is more serious, and the turnover rate can reach 20%. The cost of getting a new customer is five times that of maintaining an old customer. Therefore, it is especially important to dig out the information that has an impact on the loss from the massive customer transaction data and establish an efficient customer turnover warning system. The main reasons for customer loss are: price loss, product loss, service loss, market loss, loss of promotion, technology loss, and political loss. There are times when the price is caused by customer churn, but in fact multiple factors work together to lead to the loss of customers. For example, unrealistic profit targets, unreasonable price structure, too complicated business processes, and unreasonable organizational structure. The basic methods of maintaining customer relationships: tracking systems, product follow-up, sales expansion, maintenance access, and mechanism maintenance. Therefore, it is necessary to establish a quantitative model to reasonably predict the risk of loss of the customer group. For example: commonly used risk factors, the number and type of products held by customers, the age and gender of customers, the impact of geographic regions, the impact of product categories, the time interval of transactions, the means of promotion, and so on. Based on these factors and historical data of customer churn, the existing customers are predicted to be lost, and different maintenance methods are provided for different customer groups, thereby reducing the customer churn rate.

License: GNU General Public License v3.0