| Summary: | 碩士 === 國立臺灣大學 === 商學研究所 === 90 === With development of information technology, relationship between enterprises and customers gets complex and prompt. Amid keen competition, resources allocation decision among customers become significant. To allocate resources efficiently and cut down waste of marketing budget, customer value analysis turns to be an important tool. In this thesis, Markov chain and RFM model are integrated to calculate customer lifetime value（CLV）. Then customer lifetime value is used to allocate marketing budget and solve the problem.
The model integrates RFM model, Markov chain and discounting method to derive profit contribution of customers in every purchasing situation. In the first step RFM model is used to define customers purchasing state while transition matrix is designed to describe probabilities among purchasing states. With revenue and cost data, profit contribution of each period can be calculated. After discounted, profit contribution is accumulated to be customer lifetime value. In order to compare the original method and real data, three months of transaction data were used. Finally, customer lifetime value can be used to indicate from the empirical case study, we have the following conclusions:
1.The new model performs better than the traditional customer lifetime value estimate method, especially when the retaining rate is low.
2.Customer purchasing behavior and probability are estimated by using Markov Chain. Customer lifetime value can be used to resource allocation.
|
|---|
