| Summary: | 碩士 === 國立臺灣大學 === 國際企業學研究所 === 87 === Multi-level marketing business, 1ike Anway, Avon, Forever, Nuskin etc., has been very popular in Taiwan. Consumer is widely familiar with interpersonal selling of package products. For getting higher market share, each firm makes tremendous efforts to keep intimate relationship with their customers. Therefore, database marketing is received much attention in direct marketing industry. Through database marketing, firms can allocate their marketing more efficient and effective. However, direct marketing firms still don''t exactly know how to design suitable marketing strategy for different customers even with established customer database.
The purposes of this research are twofold. One is aimed to predict the loyalty of customers by their purchasing history of some direct-marketing firm. The other is to analysis product package by product items of this firm. The topics of thesis are:
1. Estab1ishing Hierarchical Bayes Regression Model by continuity-selling consumer database. Hierarchical Bayes Regression Model is proved more accurate than Poisson Regression Model and Negative Binomial Regression Model in predicting buying orders of each customer.
2. Reducing dimensions of product items by factor analysis in cross product database and exploring the relationship between product items and customer demographic variables. Therefore, it helps firms classify some product packages and formulate more feasible product assortment for target markets.
|
|---|
