Collection And Analysis Of Retail Information And Construction Of Marketing Business Model In Mobile Environment

With the development of mobile Internet,marketing through mobile terminals has become an important way of enterprise marketing.Mobile marketing can further subdivide customers through instant messaging and interactive communication original extensive marketing will turn to precise personalized marketing.In this paper,there are more than 5 million retail customers.They collect massive amounts of market,customer and transaction information every day,and are badly in need of information support in mobile marketing.Therefore,this paper implements the retail terminal under mobile environment information collection,building marketing business model of retail terminal based on information,combined with the practical application of the enterprise business needs,design and implementation of mobile terminal marketing management system,realize the personalized marketing of retail customers.The specific research contents are as follows:1)Data acquisition by the mobile terminal.The mobile terminal Hybird development framework is used to adapt to the continuous change of the collection task plan.Acquisition system includes basic information maintenance and updating of terminal retailers,price information collection and sales data collection,and patrol point planning for patrol salesmen.In order to improve the accuracy of terminal data and the efficiency of acquisition,this paper proposes a method based on Baidu GPS location algorithm and improved K-MEANS algorithm to optimize the data acquisition process,improve the efficiency of collection and the quality of data collection.2)Constructing the customer marketing capability grade matrix,and carrying out data mining.According to the business situation of the application enterprise,we divide customer marketing capability into customer sales ability and customer contribution degree,design the optimized random forest algorithm to classify,and design the optimized K-MEANS method to cluster.Among them,in view of the problem that the random forest can not handle the non balanced data well,the FCM_SMOTE algorithm is proposed to optimize it.For the optimization of K-MEANS algorithm,first through the factor analysis the dimensions from six dimension to four,to determine the initial clustering center of K-MEANS algorithm by using the improved shuffled frog leaping algorithm,the initial cluster center will then determine to K-MEANS algorithm,avoid the problem of local optimum,according to business needs,develop appropriate marketing strategy.3)Building the marketing business model.Marketing business model includes customer marketing ability level prediction module,formulate personalized customer order module and terminal sales information analysis module analyzing the commonly used forecasting algorithms,the GM(1,1)model is used to predict and analyze the customer marketing capability.Then,according to the results of the forecast analysis and the actual marketing situation of the customers,the individual orders are made to the customers.Through the analysis of market information,help enterprises to provide scientific basis for the development of the next step of the marketing strategy of sales.4)Constructing the good marketing business model into the system based on the actual business needs of the application of the enterprise.The mobile terminal marketing management system is designed and implemented.After a period of trial,the retail customers faced by the enterprises have been well classified,and their profits and customer satisfaction have been improved.The personalized marketing effect under the mobile environment is remarkable.