Nowadays, the multidimensional rating systems are increasingly popular in practice to help consumers evaluate multiattribute products or services. This study aims to examine the helpfulness and economic impact of multidimensional rating systems and the moderating effect of product type (ordinary, premium) from the perspective of product line. To identify the different effects of different product characteristics on consumers' behavior, multidimensional ratings and textual reviews are classified into two different sets according to the product attributes (functional, hedonic). Besides, the effects of review valence are considered simultaneously. Econometric analysis and text mining are conducted on the real data from autohome.com, a leading Chinese automobile rating platform. Results show that (a) in terms of perceived helpfulness, reviews with positive functional ratings or with negative hedonic ratings are perceived more helpful for ordinary product, whereas reviews with positive hedonic ratings are perceived more helpful for premium product. (b) Regarding product sales, the impacts of functional and hedonic ratings on product sales are inverted U‐shape. Furthermore, for ordinary product, variance of functional ratings has a negative effect on product sales, whereas hedonic ratings variance has an opposite impact. Surprisingly, ordinary product consumers tend to pay more attention to the negative hedonic reviews, whereas premium product consumers concentrate more on the negative functional reviews. In summary, various influences of multidimensional reviews and ratings across product type on review helpfulness and product sales are deeply explored in this study, suggesting that firms' online marketing strategies should be contingent on product type and characteristics.