Rough set computing method is used to deduce the uncertain and incomplete information collected from group-buying websites, the conclusion is that the more important impact on group-buying website sale is relative average original price, group-buying commodity types, group-buying activity times and buying person-time, average discount rate and hot-city number, that the more group-buying activity times and buying person-time have, the higher the sale is, however, average discount rate and the number of hot-city have small influence on group-buying website sale. As a result, compared with price, credit plays a more important role in group-buying website development. More customers should be attracted through group-buying guide websites and development strategy with different commodity guide should be implemented so that excellent service can be used to promote the loyalty of customers by group-buying website.