Journal of Systems & Management ›› 2020, Vol. 29 ›› Issue (5): 924-933.DOI: 10.3969/j.issn.1005-2542.2020.05.010

Previous Articles     Next Articles

An Empirical Study of Purchase Rate and Dropout Rate Between Mobile and PC Customers

WU Banggang, YANG Liu, CHEN Yubo   

  1. 1. School of Business, Sichuan University, Chengdu 610064, China;  2. School of Economics and Management, Tsinghua University, Beijing 100084, China
  • Online:2020-09-29 Published:2020-10-26

移动电商顾客购买行为——基于拓展BG/NBD模型的实证研究

吴邦刚1,杨柳2,陈煜波2   

  1. 1.四川大学 商学院,成都 610064;2.清华大学 经济管理学院,北京 100084
  • 作者简介:吴邦刚(1989-),男,博士,副教授。研究方向为营销模型。
  • 基金资助:
    国家自然科学基金资助项目(71532006,91746302,71991461,71902148);教育部人文社会科学重点研究基地项目(16JJD630006);教育部博士后基金资助项目(2019M660243);中央高校基本业务费资助项目(skbsh2020-25,2020自研-商学C07)

Abstract:

In the mobile Internet era, one critical question for marketers is whether the purchase behavior of mobile customers is different from that of PC customers. To address this issue, this paper extends the classical BG/NBD model in customer relationship management, and studies the impact of adopting mobile e-commerce on customer purchase rate and dropout rate. It also examines the moderating effect of customer cross-buying behavior. An empirical analysis of a sample of 19189 customers from the leading online B2C platform JD.com in China indicates that mobile customers have a higher purchase rate and a lower dropout rate after the adoption of the mobile channel, and cross-buying enhances both effects. This paper provides important managerial implications for mobile commerce and customer relationship management in the mobile internet era.

Key words: mobile commerce, purchase rate, dropout rate, cross-buying, BG/NBD model, stochastic models

摘要:

基于移动渠道的特征,拓展了客户关系管理领域具有代表性的BG/NBD模型,通过在模型中引入协变量,并考虑不同交叉购买行为下,分析移动客户和传统线上客户在购买率和流失率上的差异。进一步,以2012~2014年京东商城上19189名客户购买数据为分析单位,对拓展模型进行估计和分析。研究结果表明:移动客户的购买率要显著高于传统线上客户购买率,而且移动客户的流失率显著低于传统线上客户的流失率。同时发现,交叉购买行为对上述关系具有显著的调节作用:在有交叉购买行为的情况下,移动客户与传统线上客户无论是在购买率还是流失率上的差异要比没有交叉购买行为下的差异更显著。

关键词: 移动电子商务, 购买率, 流失率, 交叉购买行为, BG/NBD模型, 随机建模

CLC Number: