Research And Application Of Bank Customer Churn Based On Data Mining

Currently,financial system reform and internet finance innovation become a much stronger challenge to the retail business in Banks.Under the situation that the homogeneity of financial products is extremely similar,customer loyalties are in the down trend progressively,customer churn is now a severe issue which all commercial banks have to face.As the largest state-owned commercial bank,The ICBC suffered from an increasing customer churn,especially from middle and high-end customer under the strong competition with all kinds of equity banks and emerging internet financial companies.This article starts from studying the current situation of customer churn to existing issues,further to worldwide applications,together with years' working experience in bank,using theoretical study on data mining and customer relationship management and empirical research based on one bank data,applying decision tree method and logistic regression model.The article depicts the modeling of an early alert of customer churn based on SAS EM data mining laboratory.At the same time,the results of this article have been applied into the real marketing activities.Taking realistic transaction data of the high-end customers as the basis,to the top 30% customers in terms of churn rate,author will promote pertinent marketing and sales plan on one bank's accurate platform,and then summarize several revelations on retrieving loss customers for further reference to related departments.It has been proved that the loss rate of customers can be reduced for almost half.Lastly,this article investigated the strategies of retaining customers through the process of modeling and the following marketing,as well as summarized several revelations on reducing customer losing and setting up the model for further reference to related department managers.The main achievements of this article can be summed up for these three aspects:1st Set up an early warning model of customer lost.2nd Apply the model into the real marketing of retaining customers.3rd Prove the effectiveness of losing customers retaining measures on reducing rate of losing customers.