Demand effects of product similarity network in e-commerce platform

With the increasing popularity of product assortments by recommender system, it becomes increasingly important for online platform and sellers to investigate their economic impact and manage the links of products. Different from the previous product network, this study constructs and investigates the role of similarity product network from similar products’ recommender system by using data from Taobao.com. The characters of similarity product network exhibits influence on product demand. Mining of similar product’s link reveals that the more a product is being linked, the greater of the demand, the impact of product’s degree is different from type of products. In addition, the results show that the centralization of network has a negative impact on focal product’s demand. This study also examines the spillover effect of similar products’ reviews (UGC) as well as similar products’ description (MGC), especially focuses on the semantic similarity. The results reveal that the semantic similarity of recommended product’s reviews and products’ description have negative spillover effect on demand. The more similar of recommended product’s reviews and the product’s description, the stronger the effect. Specifically, similarity of MGC exhibits a stronger impact than that of UGC on focal product’s demand for search goods than for experience goods. The findings provide insights to marketing practitioners by helping understand the effects of similarity and link of product on the consumer’s decision.