The Design And Implementation Of The Precision Marketing Model For The Initial Issuance Of Credit Cards

With the continuous advancement of scientific research methods and the penetration of cutting-edge technologies in the financial field,more and more data mining technologies are beginning to be applied to various business lines of banks.According to the analysis of massive data,it provides the development of various lines of banks with supportive decision-making basis.Correspondingly,with the increasing number of credit card users,more and more commercial banks are beginning to use data mining to seize the credit card market.Due to some policy reasons,a city commercial bank started its credit card business late.For commercial banks with zero credit card customers,they should avoid blind issuance of credit cards when they first issue credit cards,which may result in poor customer experience or “empty cards” or“sleeping card”.This article designs a credit card initial issuance precision marketing model in the credit card initial issuance marketing scenario.Based on the traditional customer segmentation model theory,the design superimposed the predictive card opening model on the customer segmentation model,and implemented the model in the actual environment.This article designs a credit card initial issuance precision marketing model in the credit card initial issuance marketing scenario.Based on the traditional customer segmentation model theory,the design superimposed the predictive card opening model on the customer segmentation model,and implemented the model in the actual environment.The main work completed by the study are:First,In-depth analysis of the principles of traditional customer segmentation models,combined with the business scenarios of this topic,summarizes the limitations of traditional models in practical applications.Then extract the principle of customer segmentation based on the traditional customer segmentation model,so as to screen and combine customer attribute dimensions in the data mining data preprocessing process to combine the operation of the predictive card opening model.Second,a new model was designed—the precision marketing model for initial credit card issuance.The new model consists of two parts: predictive activation card model,analysis of customer characteristics of potential credit card demand customers through customer credit card repayment records.Based on the customer segmentation theory,the customer segmentation model is superimposed on the predicted card activation model.The model algorithm uses an election algorithm based on XGBoost and Light GBM,the final combination forms an accurate marketing model for initial credit card issuance,then further screens the customer set with new credit card needs.Finally,the implementation and application of the credit card first-issuance precision marketing model was completed,using the resources of more than 40 million active debit card users in a commercial bank to complete the analysis of the card opening needs of existing customers in the shortest time.Practice has proved that after the credit card system goes online,the output of the model helps the business personnel quickly locate the customer base.Based on this,they quickly launch marketing activities,reducing the time cost of searching for targets and increasing the success rate of credit card marketing.