Whom Whom do I Choose to Diffuse Information on Twitter? An Agent-Based Model Approach

Abstract

Social network sites such as Twitter have made easier the exchange of information about products among consumers that are users of these networks. Thereby, marketers are using word of mouth marketing campaigns to engage users of these social networks in spreading the word about their products. The choice of consumers who participate in these campaigns is key in their success. However, it is not clear which type of consumers and its combination are the most effective in reaching this aim. Unfortunately, the difficulty in getting useful data due to its noise or non-public availability do not allow researchers to appropriately apply traditional methodologies like statistics. Hence, this paper proposes an agent-based model to simulate the information diffusion through Twitter when some types of consumers participate in word of mouth marketing campaigns. This model was implemented in Java and its results show that, although hubs (users with many contacts) are better connected than opinion leaders, the latter ones reach higher information diffusion because of its influence; in addition, the use of both types of consumers in the same campaign achieves as diffusion as using just opinion leaders. Apart from results, this paper gives to future researchers a guideline to develop agent-based models that simulate different behaviors of Twitter’s users.