Investigation And Analysis Of M City Mobile Customer Churn

China’s mobile communications industry is gradually occupied by three communications providers; therefore there is more intense competition among operators. We all acknowledge the truth that the customer is the center of the market competition is leading that competition for customers is becoming the focus of the three operators, which has been bringing more serious churn problem. This phenomenon not only breaks the balance of the market, but also reduces the operators’ economic benefits to varying degrees. Almost every global telecom companies are being or is about to establish customer churn models around the world, because enterprises without establishing Loss Model lack market competitiveness will even lose foothold in the industry.For the purposing of explore the reason of the communications industry’s loss of customers, in this paper, the author aimed at the problem of customer loss in this field, which is based on the theory of customer relationship management and data mining theory, and then put forward the prediction of the customers loss based on Logistic regression model. Firstly, according to the consultation with the relevant people and the experience of the past, there are 18 factors on primitive data collection that may result in loss of customers. Secondly, using data samples of the survey and statistical forecasting model, the author finds out 8 key elements influencing the customer defection. Finally, the author obtains an estimated function form and test it, moreover, saving the results show that the model’s accuracy is high.On the above analysis, we draw the following conclusions:(1) the number of the main cause of customers loss in city M is eight.(2) Among the factors affecting customer churn, some play a positive role, but others play a negative role.(3) In the light of the estimated classification function, the author put the eight factors in order following the importance from the most to the least in accordance with the coefficient. The maximum coefficient is the factor called whether used other carriers or not, the minimum is total cost. To be clear, these eight factors are the main factors. It doesn’t mean that there’s nothing else or the others have no use.According to the Logistic regression model combined with the model variables and external factors, this article put forward some policies and suggestions from the perspective of mobile company that promote customer complaint management. This article takes advantage of the preliminary findings and provides my own solutions for China Mobile.