利用不同加權法評估顧客活躍性指標

Abstract

購買間隔時間是行銷領域中常常使用的變數之一，企業可以透過購買間隔時間的變化來衡量顧客長時間的消費趨勢，也就是判斷顧客的活躍性，而過往針對顧客活躍性的分析最常使用的就是CAI指標，但其中計算加權時的權數給予在高購買頻率時會顯得不合理且不適當，因此本文的研究引入了指數移動平均(EMA)這個方法當作新的加權模型，並與原始CAI的結果作比對，確實發現了高購買頻率下原始CAI的不靈敏，而EMA CAI在此狀況下則是的表現較佳的。 傾聽與關心這些高價值顧客的意見是企業差異化策略的關鍵之一，使用本文所推薦的EMA CAI便能更清楚觀察到這些原本購買次數多、忠誠度高的消費族群近期的活躍度是否明顯下降，才能讓企業的顧客流失率下降並維持較高的收益。

Inter-Purchase time is one of the variables commonly used in the marketing field. Enterprises can measure the long-term consumption trend of customers through the change of purchase interval. That is, judge the customer's activity, and we usually use CAI to evaluate a customer is active or inactive. Though the weight given when calculating CAI is unreasonable and inappropriate at high purchase frequency. Therefore, the research in this paper introduces the exponential moving average (EMA) method as a new weighted model. Compare these two CAIs, we observe the original CAI is insensitive at high purchase frequencies, while the EMA CAI performed better under this condition. Listening and caring for these high-value customers is one of the keys to differentiation strategy. Using the EMA CAI recommended in this paper, it is more clear that the recent activity of these consumer groups with many purchases and high loyalty is obvious. Enterprises have a decline in the company's customer churn rate and maintain high returns.