顧客活躍性機率與顧客終身價值分析

Abstract

本研究主要研究目標為幫助管理者辨別出正確的現有顧客以及潛在顧客。藉由資料探勘採集顧客購買行為資訊，並透過Schmittlein等人於1987年所提出的Pareto/NBD模型以及Fader等人於2005年提出的顧客終生價值模型，分別計算每位顧客的活躍性機率與其終身價值；再以變異數分析(ANOVA)，建構出具獲利性顧客輪廓(Profitable customer profile)，本研究採用網路光碟零售商X-Company於2006年顧客交易資料庫為實證研究資料庫。計算每位顧客的活躍性機率與終身價值可以協助管理者進行有效的資源配置，提高客製化與直效行銷之效果；建立獲利性顧客輪廓可以協助管理者獲致具獲利性的潛在顧客的基準以及開發具獲利性的新市場等。研究結果顯示X-Company的具獲利性顧客被分為四個優先順序等級，管理者可以依據該獲利性顧客輪廓進行行銷資源配置以及制定行銷策略，做為獲致具獲利性的新顧客之基準。

The major objective of this research is to assist marketing managers in identifying and learning from their “right customers” that helps managers to serve high-value customers, and create successful marketing programs targeted at the prospected customers. Through data mining of individual transaction information, and Pareto/NBD model proposed by Schmittlein et al. (1987) and customer lifetime value model proposed by Fader et al. (2005) are applied to understand customer active probability and customer lifetime value. Then apply simple one-way ANOVA to construct profitable profile that improves the effect of acquiring new customer. The data on the purchasing of VCDs at the online retailer X-Company is used as empirical analysis. The research result demonstrates that there are four priority ranks of X-Company’s customer. Through the profitable customer profile, managers can observe customer’s purchase behavior to choose the prospects that best match the profile of company’s profitable customers.