分量迴歸模型之顧客價值分析-以油品消費為例

Abstract

本研究以汽、機車顧客之加油消費行為做為資料庫行銷之實證研究資料，針對資料庫內之顧客個人資料以及交易資料，採用分量迴歸之統計方法探討顧客價值分析：針對平均每日交易金額、平均每日交易次數以及活躍性三項顧客價值衡量指標，將市場分為高、中、低顧客價值群三個區隔市場，藉由迴歸模型的估計 結果描繪各別市場內之消費者人口統計以及消費特性。 首先在平均每日交易金額和交易次數價值分析中，本研究分別以分散到各顧客消費期間之平均每日交易金額和次數做為依變數，資料庫內之人口統計變數為自變數，探討各別自變數在平均每日交易金額和次數之條件分配各分量下之邊際估計結果。並依據依變數顧客價值分配之分量，將顧客分為高、中、低三個市場區隔，藉由描述不同市場區隔內之消費者特性以構思可行之行銷策略。 此外，在活躍性價值分析中，以最大概似估計法以及加權最大概似估計法計算出消費者之平均購買期間和加權平均購買期間，並進而將兩項購買期間平均數以比例的方式呈現出活躍性指標。接著以活躍性指標做為顧客價值衡量指標，同前述之分析方式，得出不同市場區隔下之消費者特性分析以及可實行之行銷策略。 最後說明該研究使用之資料庫可能產生之限制，並針對該限制提出未來研究可能進行方向之建議。

The main purpose of this research is to help the company establish a better understanding of its customers, figure out the heterogeneity of each one and adopt one-to-one marketing strategy rather than mass marketing strategy. Thus, we use customers’ past transaction data and apply them to Quantile Regression model to identify customers of different segmentations more precisely. First of all, we use three customer value criteria: daily transaction money, daily transaction frequency and the ration evaluating customers’ consumption variation of time. And we apply both OLS and Quantile Regression Model to the customer data to compare the outcomes between two statistics techniques. Besides, we use Quantile Regression Model to segment customers to three groups: (1) quantile0.1~0.3: low-value customers (2) quantile0.4~0.6: middle-value customers (3) quantile0.7~0.9: high-value customers And than we use the estimating outcomes from Quantile Regression Model to describe customer characteristics of three groups. In the end, we point out the possible restrictions from the database and make some suggestions for future research.