A Research On Precision Marketing About Turning Single Broadband Into Multi-Services Combo Based On Data Mining Of ZJ Telecom

With the basic telecommunication market tends to be saturated,the competition within the telecom industry is increasingly fierce,as of the national policy is constantly inclined to the consumers and the Internet industry is rising.Facing pressure from increasing cost and declining income,previous rough operation mode is no longer adapted to the new era for telecom operators,and it puts forward higher requirements for delicacy management.It is an urgent time to telecom operators to launch a business mode transformation as well as looking for new business growth opportunities.At the same time,big data is changing the world gradually,it is an opportunity as well as a challenge for telecom operators.There are major issues with current telecom operators:how to make good use of its data resources advantages,how to combine with the edged technology of this industry,breaking the traditional operation mode intends to upgrade business,to help delicacy management and to improve business operation efficiency.This paper takes ZJ Telecom as the research object and combines the data science theory with the practical business by using the research methods of literature research,quantitative analysis,experience summary and so on.Taking the precise marketing of single broadband consumer turn into multi-services combo customer as the subject,this paper uses the data mining tool SPSS Modeler to construct the categorizing prediction model.As the recognition rate of target customers increases,it can use the limited marketing resources more efficiently.Meanwhile,based on the existing customer characteristic data,it divides these customers into specific groups,then designs marketing programs for each customer with experience,which support the development of practical business and realize the value of the data.This article introduces theory and knowledge of data mining and categorizing model,based on practical business,as well as illustrate business goal of predictive model on turning single broadband into multi-services combo.By deeply understanding business and service,it has been undertook to design end customers data table,and clarify data period then observe data group,by the way of data regulation,data transformation,data balance and so on in order to initially process data,combining with various categorizing algorithm such as C5.0 and random forest to set up predictive model separately.After repeated adjustment and evaluation,it could be accomplished to confirm the best model.According to output from predictive mode,with practical experience and segmenting end customers by characteristic,service set recommendation strategy,promotion tactics,promotion campaign and advertising channels could be well designed,for final goal achievement of supporting real business development.