The aim of this study is to explore bidding preference of members in virtual community by observing the bidding timing on different price of product and types pf social network. Data are collected from the community of "bidding shoes!" of Facebook. Totally, 300 bidding transaction records are collected. Results show that bidders have different preference on bidding timing in different price of product. Furthermore, the study observes five types of social network of bidders: simple bidding structure, mutual bidding structure, sequential bidding structure, multi-bid structure, complex multi-bid structure and intervention bidding structure according to their characteristics. The most observed type is the intervention bidding structure. And a further study is then conducted to identify bidding behavior on the intervention bidding structure. Results are expected as a reference for sellers and community managers to achieve appropriate auction opportunities.