营销学系 | 数智时代营销洞察系列学术讲座(一)

Store Visits, Locations, and Customer Perceptions: Market Structure Analysis with Customer Trajectories in Shopping Malls

2026年4月27日（周一）13:30-16:30

宁远楼213会议室

姚岱博士是香港理工大学商学院营销学副教授。他在清华大学、新加坡管理大学和欧洲工商管理学院(INSEAD)分别获得计算机本科、运营管理硕士、以及市场营销学博士学位。他的研究和实践主要基于统计学、人工智能、神经科学、以及经济学等多学科知识融合，并聚焦在高新技术赋能的各类商业场景里的用户和企业的行为分析。他在世界顶级管理和营销期刊发表近十篇论文，包括《营销科学》(MKSC)，《管理科学》(MNSC)，《营销研究期刊》(JMR)，《消费者研究期刊》(JCR)，《生产与运营管理期刊》(POMS)，《营销科学学院期刊》(JAMS)，并担任这些期刊的匿名评审以及《商业研究杂志》(JBR)副编辑。

This paper conducts market structure analysis in retail agglomerations using unique trajectory data from 10 million customers across two shopping malls. We use a scalable deep embedding method to construct “store co-visit similarity” based on co-visitation patterns. This metric captures both consumer brand preferences and the impact of store locations, summarizing realistic inter-brand relationships for brick-and-mortar stores. We augment the trajectory data with consumer surveys and mall layouts, allowing us to decompose the store co-visit similarity into stated brand preferences (28%), store locations (22%), and other unobservables (50%). This decomposition suggests the substantial role of factors beyond stated preferences and location in shaping inter-brand relationships. We demonstrate our measure’s effectiveness in capturing substitution patterns through two applications. First, we show how our measure captures consumer substitution in response to temporary store closures. Second, we conduct a counterfactual analysis of changing store locations, showing that co-visit similarity better predicts transition probabilities between stores compared to a model that simply combines perception and location data.

图片、文源|国际商学院营销学系

美编|国际商学院宣传中心