Construction And Application Of Tobacco Precision Marketing System Based On Big Data

With the increasingly fierce competition in the marketing and the gradual acceleration of market segmentation,each marketing industry is faced with such problems as effective customer management,brand sales region selection and product delivery strategy.How to use the existing resources and technologies to achieve precision marketing to solve the above problems has become an urgent problem for all marketing industries.In recent years,big data has gradually developed and successfully applied in many fields.This thesis takes the tobacco marketing industry as an example,combing with big data analysis and extracting consumption information in the marketing process to construct a precise marketing system based on big data.To help a tobacco company manage customers,grasp the risk of customer lose,select brand sales regions and delivery strategies,to some extent,solve the difficulties encountered in the management of the tobacco industry,and play an auxiliary role in the management of the tobacco enterprise.The main contents of this thesis are as follows:1.Process and analyze the tobacco marketing data based on big data technology.First of all,We integrated the enterprise’s marketing data distributed in different systems,established the enterprise’s own marketing data resource pool,and conducted data cleaning and conversion,integration of internal and external marketing data,correction processing operations,form a usable data set through some operations such as storing,cleaning,transforming;Secondly,design the data warehouse model according to the business analysis theme by the help of OLAP analysis model(such as retail,sales,inventory or planning);Finally,the obtained massive data is analyzed and processed in depth,and the sales forecast model was built through decision tree regression,XGBoost,Light GBM and other algorithms to find the hidden value behind the data,which is presented in the form of charts and other intuitive forms.2.A tobacco precision marketing system based on big data was built,including the overall architecture design of the precision marketing system,customer analysis,brand sales region analysis and delivery strategy.This thesis analyzes the reasons of customer loss with the purchases number and average consumption,classifies customers with the customer segmentation indicators to clarify the types of customers.K-means clustering is used to analyze the sales status of a brand in each sales area,and find out the potential influencing factors;Improved CFSFDP algorithm is used to analyze the customer value,so as to clarify customer consumption habits.The customer segmentation matrix is introduced to develop marketing strategies for customers of different values.BCG matrix,as an evaluation criterion is adopted to establish market scoring rules,analyze brand market status,so as to adjust the market strategy to achieve precise delivery.This study improve the big data processing efficiency and marketing efficiency of tobacco companies,enhance the adaptability of the system to decision support,as well as the market positioning and sales decisions for new product research and development,so as to achieve precise marketing for customers.