Abstract: Whether omnichannel retailing can improve the longterm performance is the current focus of retail enterprises This paper uses customer lifetime value to evaluate longtime performance and examine the relationship and mechanism between omnichannel retailing and customer lifetime value，by using big data and machine learning approach The study finds that omnichannel retailing can effectively improve customer lifetime value The process is moderated by heterogeneous consumer demand，such as consumers preference for commodity diversity，fulfillment cost sensitivity and storage cost sensitivity This paper argues that retail enterprises should make firm strategic decisions on omnichannel transformation，and provide consumers with a combination of omnichannel goods and services in a targeted manner based on the heterogeneous needs of target consumers to improve omnichannel efficiency

Key words: omnichannel retail, customer lifetime value, machine learning, consumer heterogeneous needs

摘要： 全渠道零售能否提高企业的长期绩效是数字化时代零售企业关注的重点。本文采用了机器学习方法，利用消费者大数据训练客户终身价值的预测模型，并对全渠道零售和客户终身价值的关系进行回归分析。分析结果显示，全渠道零售能够有效提高客户终身价值。进一步研究发现，这一过程受到消费者异质性需求的影响，全渠道转型对客户终身价值的影响会由于消费者商品多样性偏好、履约成本敏感度和仓储成本敏感度的不同而存在显著差异。据此，本文认为零售企业应当坚定全渠道转型的战略决策，立足于目标消费者的异质性需求，有针对性地向消费者提供全渠道商品和服务组合，提高全渠道效率。

关键词: 全渠道零售, 客户终身价值, 机器学习, 消费者异质性需求