Multi-relationship Based Social Network Formation In Social Commerce And Its Impact On Transactions

This research conducted three studies to explore the multi-relation based virtual resources accumulation and the correspondant marketing strategies under the context of social commerce. In the specific researches, we choose the transactional community which features the commercial relationship among its community users as the research object to explore the multi-relationship among users and the economic values of these social relations so we can finally help enterprises design their marketing strategies to use the virtual social resources. The first part of this research begins with the drivers of social network evolution and continues with the exploration of the formation and evolution of the multi-relationships in the social commerce community under different drivers. The results show that the users in transactional community as buyers and sellers would build ties with others under different drivers. From the perspective of buyers, they are more likely affected by observational learning when they build ties with other buyers and they are more likely affected by both observational learning and contagion when they build ties with other buyers and sellers; from the perspective of sellers, homophily, reciprocity and structural equivalence are the major drivers when they build ties with both buyers and other sellers. The second part of the research explores the dynamic corelations between the formation of relationship and the transaction records of the sellers using the vector auto regression model. The results show that the incoming ties form both buyers and sellers would increase the sales of the sellers significantly, and this effect would climb up to the peek after four days of the relationship formation. In contrast, the out-going ties from both buyers and sellers to buyers would not increase the sales of sellers significantly. Besides, the results from the impulse-response function show that different types of relationships would have different effect on the sales in the view of both short term and long term. Specifically, the ties between sellers would have the strongest effect in the long term (1.306) and its cumulative effects are four times higher than the effects of ties formation between buyers and sellers (0.327). In contrast, the ties formation between buyers and sellers and within buyers themselves would not have significant effect on the sales of sellers. In the last section, we focus on the strategy design of viral marketing in transational community and study the approaches to promote the marketing performance based on the virtual social resources we have previously studied. The results from Agent Based Model Simulations (ABMS) show that there is significant difference for certain information to diffuse under the mechanisms of social influence versus homophily. The difference is still significant after we control for the network density as well as other variables. The difference keeps significant when we randomly choose the samples from the real-world network data. Specifically, the mechanism of homophily can increase the range of diffusion right after the initiation of the viral marketing strategy, but as the time goes by, the mechanism of homophily can not keep the increasment in a long term due to the limitation of community activities. On the other hand, the mechanism of social influence would not be under the constrant of community activities, which can diffuse the information into wider range of potential consumers. This study also focuses on the targeting of seeding consumers, the results show that the hubs in an inter-personal network are not the best choice of seeding targets in transactional community. To optimize the viral marketing strategy performance, we can choose the users as hubs in both inter-personal network and affiliation network. In the last, we empirically test each mechanism in the process of diffusion and get the conclusion that the mechanism of homophily has more significant effect than social influence during the process of viral marketing.