不同交易情景下客户终身价值（CLV）建模研究综述--1985年到2011年.doc

  不同交易情景下客户终身价值（CLV）建模 研究综述—1985 年到 2011 年# ， ，马君* 5 10 15 20 25 30 35 （北京邮电大学经济管理学院，北京 100876） 摘要：客户终身价值（Customer Lifetime Value, CLV）的建模从 1985 年开始, 引起学术界和 实业界越来越多的关注, CLV 预测的建模方法得到了很大丰富, 但是这些学术研究的成果并 没有在实践中得到广泛应用, 究其原因, 模型的适用条件探讨不足是影响模型应用的重要因 素之一。将交易情景分为非契约型、半契约型和契约型, 将客户的交易行为分为连续交易和 离散交易, 从而将 CLV 建模领域划分为六个象限, 对每个象限内的 CLV 建模方法进行综述, 总结出不同交易情景下的 CLV 建模的关键预测变量和计算公式。根据综述可以得出, 目前 在非契约情景下的研究最为丰硕, 契约交易情景下的 CLV 建模也受到越来越来多的关注, 而半契约情景下的研究几乎空白。研究结论指出了三种情景下未来的研究方向, 其中较为详 细的勾勒出半契约交易情景下 CLV 建模的未来研究思路。 关键词：客户终身价值；契约型交易；非契约交易；半契约交易 中图分类号：F270 The Review of Customer Lifetime Value (CLV) Modeling in Different Transaction Contexts---from 1985 to 2011 WANG Zelong, QI Jiayin, MA Jun (School of Economics and Management, Beijing University of Posts and Telecommunications, Beijing 100876) Abstract: Customer lifetime value (CLV) modeling has attracted increasing attention in the fields of academy and industry since 1985. There are some academic results for CLV prediction, but most of the academic research has not been applied by industry. One of the important reasons is the insufficient exploration for the using condition of CLV models. The transaction contexts between customers and company are divided into non-contractual transaction context, semi-contractual transaction context, and contractual transaction context, and then, the transaction behaviors of customers are separated into the continuous and the discrete ones. So we classify the CLV modeling methods according to the six above-mentioned quadrants. What’s more, we review the CLV modeling methods within each quadrant to summarize the key predictor variables and the formulas in different transaction contexts. Therefore, we conclude that the research of CLV modeling is fruitful in the non-contractual transaction context and is increasingly noticed in the contractual one. However, in the semi-contractual transaction context, it’s nearly blank. Based on these findings, we point out the future researches of the three transaction