Researchon Sales Strategy Based On Data Mining Algorithm

Along with the progress of society,the rapid development of productivity level and information technology,our daily life has been affected and changed all the time.A variety of new technologies and new products emerge in our life continually and there are changes in people's shopping way that people will more turn daily shopping to online shopping,because online shopping is not subject to the time and geographical constraints,and it can be carried out anywhere and anytime.At the same time,this change also brings us a challenge.We are faced with a larger choice space,and we have to spend more time on choosing the products we need.For enterprises,the emergence of this trend is both a challenge and an opportunity.Rivals enterprises faced are everywhere.They cannot just rely on their products to retain and attract customers,they need to focus on customers' personalized demand,and they should capture what customers need in time to perform appropriate sales strategies timely.Only by this way can they win more customers.Current customer shopping behaviors will produce large amounts of data,which contains a lot of information that has not been fully excavated and utilized.The emergence of data mining technology allows us to mine the customer needs and shopping patterns from the customer's shopping behaviors,which companies could base on to make sales strategy accordingly.To customer,it can help them save the cost of shopping to a certain extent.To enterprises,it helps them increase product sales and improve customer stickiness,so as to achieve a win-win situation.Marketing strategy research refers to deciding how to sell goods what customer may buy to them in a proper way according to their interest features and past purchase behaviors.Sales strategy based on data mining algorithm is that mining customers' interest and demands information based on their past purchase behavior and selling goods users may be interested combined with the similarity between the goods in the shopping behavior,not in the similarity of physical properties between Commodities,that is so-called portfolio sales strategy;Or according to customer's shopping behavior finding customer clusters with similar characteristics,combined with the customer shopping behavior and physiological similarity,the so-called cross-selling strategy;also it can be according to individual users' purchase behavior in the past mining users' shopping patterns and selling products to users may buy based on the purchase mode.The first mode is based on object.The second and the third mode are all based on user's,further,the former is based on user group and the latter is based on single user.All above selling strategy algorithms mainly include association rules,clustering and sequential pattern mining algorithms.This paper bases on the association rules and clustering algorithm for discussion.Firstly it shows the research of domestic and foreign,then introduces the commonly used marketing strategy and data mining algorithms in theory,and makes combination of the two based on the theory research.It is found that the combination can be effectively applied to the sales in all walks of life.In fact,there is already such a study that used data mining algorithms to sales strategy study long time ago,but this aspect of research is just an exploratory.Also it has some of this application in businesses,but the results of the study has not been fully analyzed and applied.The paper uses part of a retailer's sales record data for empirical analysis after the theoretical exploration,fully verifying the validity of data mining algorithms in marketing strategy research.However,this study also has some shortcomings.It is not extended in field of vision of the research,and the point of view is not comprehensive and does not take into account the time factor,but it is also the continue efforts direction of the follow-up study.