Using RFM Model and Data Mining to Predict Potential Consumers and Improve Marketing Efficiency－A Case Study of Y Pharmacy

本研究根據過去藥局經營文獻中，所提出之藥局經營關鍵因素進行數據實證，以了解在過去訪談或其他質化研究中所整理之藥局經營關鍵因素是否能實際影響藥局的經營狀況。 由研究結果，利用Y藥局之POS資料實證：具有不同特性的消費族群對Y藥局而言，具有不同的顧客價值以及消費行為，實證過去文獻所提出之結論。 而針對不同客群的分群方式，本研究利用購買品類的頻率及數量，將顧客做集群分析，共分出五種類型：家庭育兒族、新婚未孕族、高齡保健族、新生兒養育族及愛美小資族。整體而言愛美小資族因為購買產品之特性，有較高的顧客終生價值，且在購買行為上，以購買多樣性的表現顯著高於其他族群，但在購買頻率及近期性表現則較差，顯示出該族群有較多的可能至其他通路購買。而連鎖藥局的主要顧客，家庭育兒族，則有最低的顧客價值，在購買行為上品類較為單一，數量零碎雖購買行為頻繁，卻未能實際創造價值。 另外針對藥局經營關鍵因素也發現，藥局在活動促銷上，主要能促進消費者的購買頻率及縮短消費者購買的近期性，整體而言是能提升短期的消費次數，但在金額上卻未有較顯著的相關性，也顯示了金額的促銷並未能持續將顧客價值提升，僅能刺激短暫的消費行為。 針對未來相關研究之建議也認為，透過資料採礦，能有效將實際銷售資料轉化為消費行為的刑為變數，為未來藥局經營做更多實際數據的驗證，改善過去多使用店長經驗及質性訪談方式衡量經營成效之狀況 。As the concept of the Customer Relationship Management (CRM) is getting more and more popular. The analysis way of data mining used not only extrapolating data but revealing the meaning of what the customer will think and what kind of customer will be the most valuable. This research focuses on the study of utilizing the model of RFM which meaning regency, frequency and monetary. The research not only using three measures but also adding the profit of each customer to find who the most valuable one is. And establish a better way to cluster the customer then finding the best marketing strategy for them. According to understand the pareto principle of the 80/20 rule, the research combine RFM model and cluster to measure the customer value. As well as analyzing the basic information of customer and the transaction data to understand the customer’s buying behavior and provide customer suitable products and services. The result showed that, the different types of pharmacy customers have different customer value which confirming the past researches. According to the research purposes we also clustering the pharmacy customer in five types which are family, newly-married, petty budget, elder and pregnancy. And the result showed that the petty budget is the most valuable cluster in all of the pharmacy customers