Maximizing Customer Equity by Segmentation (MCES): Proposing a Decision Support System Based on Modified K-Means

陳家祥(Ja-Shen Chen) ； 黃俊堯(Chun-Yao Huang) ； Russell K. H. Ching

《行銷科學學報》 1卷1期 (2005/04) Pp. 43-64

摘要

雖然顧客分群長久以來是行銷者的核心工作之一，而顧客價值評估在近幾年間也是個廣泛被研究的課題，但至今為止行銷研究文獻中這兩個關係密切的研究範疇間的鴻溝一直未見正式的處理。本文介紹一以顧客分群作為手段，以達成極大化顧客價值為目標的決策支援系統。這個決策支援系統的特點包括其將顧客價值評估的精神合理而彈性地融入一個分群模式中；其在最大化顧客價值的前提下決定分群數；該系統並且以改良的K-Means法為演算核心，減少極端樣本點在傳統K-Means中可能產生的影響。文中報告一簡單的實證結果，並討論其意涵。

關鍵字

分群 ； 決策支援系統 ； 顧客評價 ； 管理者判斷

並列摘要

As segmentation has been one of the central marketing tasks for decades and customer profitability valuation has seen wide study during the past few years, surprisingly, up to this date, there is a gap in marketing research that awaits a bridge to link up of these two important and closely related dimensions. In this paper, we introduce a decision support system with the goal of maximizing customer equity by segmentation. The decision support system introduced here is unique in that it accommodates the essence of customer profitability valuation into a segmentation scheme in a sensible and flexible manner, that it suggests the number of segments to be determined by the goal of profit maximization instead of some arbitrary numerical criterion, and that central to its technical core the outlier problem which is pervasive in cluster analysis has been addressed by a modified K-Means algorithm so that clustering can reflect the pattern of the majority of ordinary observations in a data set instead of being influe nced by a handful of outliers. It followed by an empirical study. A conclusion remark was made at the end.

並列關鍵字

segmentation ； K-means ； decision support system ； customer valuation ； managerial judgment