Churn Prediction And Retention Research Of NY Mobile Company

In the era of comprehensive digitalization,the business model of operators has undergone great changes,and the demand of customers has also changed from the single demand of "voice+message" to the diversified demand of "voice + data service".At the same time,the cross-boundary competition of Internet is accelerating,OTT business directly provides services to users,and the value space of operators is constantly compressed.In such a complex environment,the enterprise should firmly grasp the customer this most valuable resource.However,in the face of the highly saturated communication market,the new speed is obviously not as fast as before,and the loss of customers is also increasing.Compared with the operating costs of new and old customers,the inventory has become an important business philosophy of operators,and reducing the customer churn rate has become an urgent problem to be solved.The method from post-evaluation to pre-prediction can find the users with the tendency of loss in time,and the adoption of targeted retention strategies based on the actual needs of users is bound to effectively control the loss of customers.This thesis to NY mobile loss situation as the background,through the analysis of the data mining method to establish loss prediction model,this paper USES the software SPSS Modeler,using data derived from NY mobile enterprise background information stored in the database of customer behavior,C5.0 decision tree algorithm is used to generate significant variables may affect the user erosion and rule sets.On the basis of previous studies by many scholars,this study will expand the variables including data service,integrated product,competitive infection and other dimensions for analysis.At the same time,the research closely follows the business development focus of the company to draw a conclusion of the model,which provides a precise target group for the next application of loss analysis,and provides effective marketing Suggestions for the follow-up work of customer retention.Customer loss research is a series of processes,and a complete customer loss research also requires the formulation,implementation,evaluation and optimization of retention plans after finding the reasons.Combined with the previous work of customer churn analysis and prediction,the prediction results are dependent on the business operators and the system,and the model interpretation is related to the decisions and actions.At the same time,the supporting management mechanism is integrated into the prediction and recovery process of lost customers,and standardized customer loss management process is set up to control by nodes.Integrate customer retention into each link,implement to the relevant responsible person,improve retention efficiency.In the process,this paper will operate the analysis system platform function integration application in order to greatly enhance the efficiency of retention.