This paper investigates how friendship relationships act as pipes, prisms, and herding signals in a large online, peer-to-peer (P2P) lending site. By analyzing decisions of lenders, we find that friends of the borrower, especially close offline friends, act as financial pipes by lending money to the borrower. On the other hand, the prism effect of friends' endorsements via bidding on a loan negatively affects subsequent bids by third parties. However, when offline friends of a potential lender, especially close friends, place a bid, a relational herding effect occurs as potential lenders are likely to follow their offline friends with a bid.
The implementation of enterprise systems has yielded mixed and unpredictable outcomes in organizations. Although the focus of prior research has been on training and individual self-efficacy as important enablers, we examine the roles that the social network structures of employees, and the organizational units where they work, play in influencing the postimplementation success. Data were gathered across several units within a large organization: immediately after the implementation, six months after the implementation, and one year after the implementation. Social network analysis was used to understand the effects of network structures, and hierarchical linear modeling was used to capture the multilevel effects at unit and individual levels. At the unit level of analysis, we found that centralized structures inhibit implementation success. At the individual level of analysis, employees with high in-degree and betweenness centrality reported high task impact and information quality. We also found a cross-level effect such that central employees in centralized units reported implementation success. This suggests that individual-level success can occur even within a unit structure that is detrimental to unit-level success. Our research has significant implications for the implementation of enterprise systems in large organizations.