Social media, including online user-generated ratings/reviews, are transforming the way people make their decisions. In today's e-commerce, many online vendors disguise as real users to provide overly inflated ratings on their own products (self-boosting) or negative ratings on their competitors' products (bad-mouthing). The underlying assumption is that such manipulations are profitable. However, it has never been explicitly and empirically tested. This study takes the initiative to quantify the effect of rating manipulation. We find that products with different popularity, age, and existing ratings are affected by rating manipulation differently. In some cases, being opposite to vendors' expectation, self-boosting manipulation can hurt the sales of their own products; Bad-mouthing can even help the target competitors to grow on the market. The risk of shooting themselves in the foot would discourage online vendors from manipulating online ratings. Furthermore, online platforms can also rely on our findings to pay special attention to detecting ratings of some particular products, which have strong economic incentives to manipulate online ratings.