Application Analysis Of Business Intelligence In The CRM Of A Retail Enterprise

With the rapid development of information network technology,the traditional economy has changed,and new economic models have been in vogue,such as economic globalization,the internet economy.In this economic trend,enterprises of all industries are faced with a severe test fi'om the market,and some characteristics such as the globalization of markets and customers and so on.Enterprises began to change the focus from the "profit center" to the "customer-centric",at the same time, customers also change from emotional consumer era to the feelings of the consumer era.As a typical form and has a strong tradition,the retail industry believe in the philosophy of "customer is God",they emphasize the importance of customer-to-business,so they should conform to the economic development,and compliance with the purpose of customer-centric service strategy.Only in this way, the retail enterprises can grasp the trend of customer needs,strengthen the relationship between customers and enterprises,and make a long-term development.Therefore,the customer relationship management(CRM) has become the number of enterprises,especially retail enterprises' research focus.And the customer segmentation is the primary task of CRM,which can be a more precise answer to who is the real customer in enterprises,and enterprises have focus on which kind of customers they should pay more attention,and what demand is the customers actual need.So,establish a good customer segmentation model,can do an effective cluster subdivision to the customers,and make a strategy targeted on customer retention and customer development,finally,achieve the purpose of business success and expanding sales.When planners enact the marketing plan,they will realize that the behavior of customer is an important determinant of making plan,and effective customer segmentation can let planners more clearly to identify the customer's behavior,it can be said that,customer's behavior is the basic starting point of customer segmentation. Therefore,in this paper,we choose the customer segmentation method which based on customer's behavior to do research.When we do this research,we should choose some segmentation variables,this require that the selected variables are able to take into account customer value and customer relationships.But actually,such variables are difficult to obtain.In this paper,besides select some of the data variables from the actual database such as customer information and purchase data,we choose another two variables like avg_monetary and frequency which base on the value matrix.At last,in order to reflect the customer's stability,we added an additional variableexist_time, it was a difference between the last time of customer purchase and the date of membership registration.In this paper,we mainly choose the data mining technology of the business intelligence,and the priority of data mining is data preprocessing.Data preprocessing is a process which is the most time consuming and is also one of the most important parts.Data preprocessing has not the only process,but only has the only standard:in data mining stage,can facilitate the data use.Data preprocessing includes three steps, like data extraction,data cleaning and data conversion.At first,we should extract the required data variables from retail enterprises database;secondly,we do some necessary screening,cleaning to the data variables we get at the first,mainly cleaning the vacancies and noise data;last,do some conversion to the data that we get at the second step,in this step,we make the data integrated into a unified range by some simple mathematical formula or transformation curve.Finally,we can obtain standardized data variables.Next,we use the cluster analysis of data mining technology,and choose eight variables which five variables from customer basic information and purchase data, remaining three variables are new added,and then we start to construct the retail business model for customer segmentation.This article is based on cluster analysis of the k-means algorithm which is one of the most commonly used cluster analysis algorithm.But at the beginning of the model running,needs to pre-set the number of categories,it means that planners want to divide the number of customers into how many categories.In some cases,determine the number of categories is very simple, but if you want to let all the variables are working,and make the segmentation model more effective,it is difficult to make sure the number of categories.Therefore,we wanted to divide the cluster analysis into two steps:the first step is use the one-way ANOVA method to determine the number of categories;the second step is use number of categories which have determine at the first step to run the model,and at last get a better segmentation results.In order to get the effective analysis of customer behavior patterns,we should extract all the features of each type that include customer characteristics and purchase behavior characteristics,and do an analysis and summary with these features for making an effective marketing and service planning for each type.At the last of this paper,we mainly have a comparative analysis of the results,it aims at the features of each group,finally,get a better marketing services programs for each group.Those programs not only can help to improve the relevance and effectiveness of marketing activities and service quality,but also contribute.to the implementation of a good customer relationship management.It can let the enterprise have a better understanding of each category and even the needs of each member,and make the maximum extent possible to meet the needs of member,in this way,enterprise can improve member loyalty,improve efficiency and effectiveness of corporate profits.In this paper,we want to use an example of business intelligence in a retail enterprise in empirical analyze,to illustrate the business intelligence has the capability to change the customer data into knowledge and information.And the business intelligence is not only a powerful way of retail enterprise development,but also a necessary skill of retail business survival.