Digital lead generation platforms: Rightsizing the seller base
研究了数字潜在客户生成平台（DLGP）中卖家基数规模对用户点击倾向的影响，发现两者呈倒U型关系，且不同产品类别差异显著，为平台管理者和卖家优化参与时机提供了数据驱动建议。
This article introduces Digital Lead Generation Platforms (DLGPs), an increasingly popular way to allow users to explore products from multiple retailers. Despite their growing influence, little is known about how DLGPs can manage their effectiveness or profitability. Here, we discuss their distinctive and salient aspects relative to other types of digital retailing, and explore data-centric methods to better manage the size of their seller base. Specifically, using a rich proprietary dataset, we examine drivers of user click propensity (UCP), focusing on a key issue for platform managers: is consumer response better when there are “endless aisles,” or should the number of sellers active at a given time (the “base of sellers”) be somehow limited in a category-specific manner? Based on a flexible nonparametric model, our results suggest that base of sellers has an inverted-U relationship with UCP, with potentially severe consequences for seller underpopulation, one that is masked when endogeneity is not corrected for. Intriguingly, we do not find such an effect for the number of offers, which might be expected based on “overchoice”. Although this general shape for base of sellers is apparent in all 10 product categories studied, there is substantial variation in how often each is suitably populated, with Cars over 60 % of the time and Mobile Accessories only 5 %. These findings have important implications for both DLGP managers and sellers: platform operators can enhance their revenue potential by “rightsizing” their seller base, while sellers may be able to improve clickthrough rates by timing their involvement based on contemporaneous competition in their particular categories.