The widespread use of Internet accelerates the rapid development of business to customer electronic commerce. To reduce information overload and help their customers to make better purchase decisions, e-commerce websites are beginning to use online recommendations. This paper compares the effectiveness of three types of online recommendations, the personalized recommendation, best sellers, and consumers reviews, which are widely used in e-commerce. This research used a laboratory experiment combined with a questionnaire. This paper also establishes an integrated model of the facts that influence recommendation effectiveness.