Abstract:
Peer influence through word-of-mouth (WOM) plays an important role in many information systems but identification of causal effects is challenging. We identify causal WOM effects in the empirical setting of game adoption in a social network for gamers by exploiting differences in individuals' networks. Friends of friends do not directly influence a focal user, so we use their characteristics to instrument for behavior of the focal user's friends. We go beyond demonstrating a large and highly significant WOM effect and also assess moderating factors of the strength of the effect on the sender and receiver side. We find that users with the most influence on others tend to be better gamers, have larger social networks, but spend less time playing. Interestingly, these are also the users who are least susceptible to WOM effects.
