The Effect Of Social Influence On Consumer Payment Behavior In Virtual Worlds

With the development of Internet, people are getting involved more and more in virtual worlds. As an application based on computer technology and Internet technology, virtual worlds enables individuals to interact with others via Internet without geographical restrictions. Social media is a typical virtual world application, e.g. We Chat, microblogging, BBS, etc. Another typical virtual world application is social gaming, like World of Warcraft, the Legend of Mir, Fantasy Westward Journey, and so on. These virual worlds applications attract numerous of users, especially after the rapidly growth of mobile smart devices. However, there is one serious problem with the development of virtual worlds. Virtual worlds usually applied freemium business model, i.e. free for basic service, and charge for advanced functions. Unlike traditional subscription business model, the freemium business model brings the virual worlds to a struggling situation: while the number of users and active users are surprisingly large, virtual worlds still cannot generate profit for the company, as the ratio of payment user is quite small. As a result, how to engage users to pay becomes a problem need to be solved. This paper is trying to answer this question from the perspective of social influence.Firstly, we constructed a simultanoues equation model using data collected from Twitter, and explored the relationship between users’ virtual world social influence power and social influence power outside the virtual world. The results indicate that social influence power in outside word is a source of virual world social influence power, and vice versa. Meanwhile, activities in the virtual world would also contribute to users’ virtual world social influence power, and this relationship is moderated by users’ outside world social influence power.Then, we explored the different impacts of different types of social influence generated by different social networks in the virtual worlds using the dataset of a social game. We introduced and extended the two-regime network autocorrelation model to investigate the different impacts of different social influence. The results shows: new friend has more influential power than old friend; structural equivalent effect is significantly larger than cohesion effect of both new friend and old friend; sharing common rivals would lower the correlationship of two friends; equality of friends’ wealth represented by Gini index is negatively correlated with players’ payment willingness.Further, we investigated the different effect of social influenc coming from different sources. We identified whether user’s friend is his friend outside the virtual world, and then performed a Tobit regression. According to the result, players are more likely to be influenced by friends known in the virtual world, but not friends known in the outside world.At last, we identified the role of hardcore player in the development of virtual market in the virtual world. The results suggest that the emergence of hardcore user would promoted the development of virtual market for both market total trading amount and market total trading frequency. However, they would finally harm the virtual market if there are too many hardcore users, as too many hardcore users make casual users feel boring in using virtual worlds.Using dataset of a social game, which is a typical virtual world application, this paper validates social influence theory in the context of virtual worlds. Meanwhile, this study contributes to the theory by exploring different effect generated by different types of social influences from both micro level and macro level. Moreover, this paper proposes social influence power building framework in virual worlds by using dataset of microblogging, which is another typical virtual world application, and it enriches social influence theory. Our results have significant managerial implications. Companies who are marketing in the virtual worlds would benefited by knowing how to increasing their influence power and how to choose potentially influential posters. Companies who are running the virtual worlds would get ideas from this paper to design a better virtual world and make better marketing plans to engage users to pay. Also, this paper sheds lights on gamification task designing for the crowdsouring companies.