【公开学术报告】Identifying the Agglomeration and Dispersion Forces in E-marketplaces Utilizing Consumer Clickstream Data

Guest Speaker:    Dr. Xueli ZHANG (Nanyang Technological University)

Time/Date:          9:30 am Monday, Oct. 12, 2020 (Beijing Time)

Length:                 1.5 hours

Meeting ID:         674 027 22562

Password:            931394

Click Link:            https://zoom.com.cn/j/67402722562

Abstract

Two opposite forces simultaneously exist among firms in marketplaces. One is the (positive) agglomeration force, which is captured by demand-side externalities, making firms tend to cluster together. The other one is the (negative) dispersion force, which is captured by the substitution (business-stealing) effect, hindering firms from co-locating in the same marketplace. Separate identification of these two forces empirically using observational data requires complete footprints of an individual’s shopping journey. In the brick-and-mortar space, such footprints are not easily traceable. Hence, the extant literature primarily focuses on the net effect or the sum of the agglomeration and dispersion forces. This paper develops an empirical framework to separately identify these two forces under the context of a cashback website. Specifically, I classify consumer traffic into four types: initial traffic, spillover traffic, diverted traffic, and return traffic, and the two effects could be separately identified by seven parameters: initial visiting probabilities; three buying probabilities of the initial traffic, spillover traffic, and return traffic; three traffic transferring proportions of the diverted traffic in the initial traffic, the spillover traffic in others initial traffic, and the return traffic in the diverted traffic, respectively. Furthermore, I examine the effect of cashback rates and their interaction effect with merchant categories and weekends on the seven parameters. The results show that, on average, cashback rates are positively related to three buying probabilities and the transferring proportions of the spillover and return traffic, but negatively related to the proportions of the diverted traffic. Besides, in general, cashback rates have a larger effect in the categories of “Beauty Vitamins & Supplements” and “Fashion”, and a smaller effect in the category of “Dining & Food Delivery”.

Keywords: Agglomeration and dispersion forces; cashback websites; clickstream data.