Doing what others do, herding behaviour has been shown to exist in various situations. Facing the information asymmetry problems in online shopping, customers are likely to purchase what others have purchased, which would make good retailers even better. However, because of the differences in customer preference and the intrinsic needs to be different from others, customers may not follow other’s decisions. With a panel data set collected from taobao.com, this study empirically examines whether herding effect exists in retailing market and how herding effect varies across product categories if it exists. This study extends the understanding of herding to an e-commerce context. By empirically showing that retailers that have a higher historical sale sell more in the future and this effect differs with different product categories, this study has important managerial implications for both online retailers and e-commerce platforms.