With the increasing popularity of product assortments by recommendation system, how online seller manage the links of products seems important. As the similar products that are viewed together in product’s search pages, we study the economic effects of the similar products network. In this study, we construct the similarity network by using data from Taobao.com. We apply Cross-sectional data models to examine the spillover effect of reviews from the similar product, additionally how characters of network moderate this effect for three product categories. We find that the semantic similarity of similar product’s review has a negative impact on focal product’s sale. The more similar in context, the more substitutable of the product. Furthermore, the volume and price of similar product also have influence on the demand. From the view of network characteristics, we find that the in-degree of a product, clustering coefficient and density of the incoming network also has an impact on the demand.