近年來企業經營的焦點在於吸引高貢獻價值的顧客,因此評估顧客價值高低的分析模型亦受到研究重視。本研究以顧客歷史交易資料進行顧客價值分析,利用顧客現值、潛在價值與購買潛力三個指標作為顧客區隔的依據。首先根據顧客交易資料的特性,提出顧客價值分析的流程與評估架構,此架構混合了資料探勘中的關聯規則與資料庫行銷中的RFM模型,分別推估顧客現在的價值、未來可能的潛在價值與顧客再度消費的可能性。除此之外並依照上述三個指標的高低作區隔化分析,探討每個群體的特性,作為行銷策略之參考。最後針對不同群集研擬客製化的行銷策略,以期藉此增進企業對顧客的吸引力,達到提高收益的目的。
In recent years, attracting customers with high contribution value has become the focus of corporate administration. Therefore, many studies are concerned with customer value analysis model. This research utilizes past customers' transaction data to analyze customer value and make customer segmentation by using Current Value, Potential Value and Purchase Potential. First, we propose an analysis model based on the characteristics of transaction data, and this analysis model combines association rules and RFM Model to estimate customers’ current value, potential value and forecast the probability of customers’ returns. Second, we segment customers according to the three above mention of indexes, and attempt to find customers' characteristics from each customer segment to design marketing strategies. Finally, we develop several customization marketing strategies for different segments, and we hope these strategies may increase corporate profits.