- 學位論文
信用卡顧客價值之分析：分量迴歸模型之應用
Credit Card Customer Value Analysis: An Application of Quantile Regression
若您是本文的作者，可授權文章由華藝線上圖書館中協助推廣。
摘要
運用「顧客關係管理」來維提升顧客價值與忠誠度已經成為目前企業的最主要課題。對企業而言，最理想之狀態是實施一對一行銷，針對不同客戶之異質性提供滿足其需求之產品。然而，由於實際上企業之資源有限，一對一行銷有執行上之困難，因此可行的辦法是透過區隔出對企業來說具有高價值之顧客，並優先將資源運用在維持這群高價值顧客身上，以期能夠提高該群顧客之終身價值，使企業獲利最大化。藉由資料庫行銷的方法，市場區隔變數為由下而上（Bottom-up）的行為變數，更能夠精確且合理的辨認出具有實質意義之目標市場，使企業在資源運用上能更有效率。 國內信用卡流通卡數自1991年的九十二萬開始成長，在2005年時達到頂峰，約有四千五百多萬張。然而由於當年底卡債風暴之影響，使得銀行大幅認列呆帳造成虧損，同時造成隔年流通卡量與簽帳金額銳減。在獲利空間壓縮的情況下，核卡條件趨嚴，也使卡友福利縮減，造成剪卡量持續增加，至2009年3月，有效卡數佔流通卡數的57%，足見使用率之低。可見超額發開發新客戶，並沒有辦法提高消費者的使用率。因此，在現在人手多卡的時代，對於發卡銀行而言，應思考如何從現有的卡友資料庫中，進行消費者行為特性與顧客價值分析，區分高低價值顧客群。 本研究嘗試以信用卡持卡人之交易行為作為資料庫行銷的實證研究對象，透過資料庫中每位持卡人的歷史交易記錄紀錄，包括交易日期、每筆交易金額、交易類別等，以及人口統計變數資料，包括顧客編號、性別、年齡、婚姻、職業、教育程度與居住地區等，來進行顧客價值分析。本研究運用分量迴歸模型作為分析工具，並分別就交易總金額、單筆平均金額、交易類別金額比例與購買期間活躍性四大構面進行分析，以期能獲得普通最小平方法結果之差異，作為區分高低價值客戶之基礎。
並列摘要
Customer Relationship Management（CRM） is a very important issue for companies nowadays to increase customer value and maintain customer loyalty. Though one-to-one marketing is the most ideal method to provide specific product or service for each customer by their heterogeneities, companies are urgent to distinguish the “most valuable customers” efficiently due to the limited resource. Thus, the preferences and needs of “high-valued customers” are the goals needed to be achieved. Consequently, companies can improve their profit and get sustainable growth by maximizing customer’s lifetime value. By database marketing, consumer’s behaviors are exactly the variables of market segmentations. In the mean time, this method could sharp the ability to find out the target market more efficiently. The main purpose of this research is to help companies to establish a better clear understanding of their customers. We use customers’ past consumption data and database marketing analysis techniques to identify each customer more precisely. This research uses quantile regression as the main analysis model to analyze customer value, and 4 variables, including “the total transactions amount”, “the average transactions amount”, “the percentage of each transaction category”, and “the active of each customer” are used to be the dependent variables.
參考文獻
廖仲仁與張金鶚（2005），「不對稱的仲介服務價格效果：分量迴歸法之檢驗」，都市與計劃期刊，第三十三卷第一期，pp.1~16 。
廖怡芳（2004），「資料庫行銷之應用—以百貨業為例」，國立臺灣大學國際企業學研究所碩士論文。
Awh, R. Y. and D. Water (1974), "A Discriminant Analysis of Economic, emographic, and Attitude Characteristic of Bank charge-card Holders: A case study," Journal of Finance, June, 973-980.
Bassett, G. and R. Koenker (1978), "Asymptotic theory of least absolute error regression," Journal of the American Statistical Association, 73, 618-622.
Berger, Paul D., Ruth N. Bolton, Douglas Bowman, Elten Briggs, V. Kumar, A. Parasuraman, and Creed Terry (2002), "Marketing Actions and the Value of Customer Assets: A Framework of Customer Assets Management," Journal of Service Research, 5(1), 39-54.