An Analysis of Word-of-Mouth Effects on Social Networks
https://computationalsocialscience.org/wp-content/uploads/2011/10/kurahashi-saito-final.pdf
This link is a research paper done on the neighbors’ effect on a consumer’s decision in regards to innovation. The paper examines two effects, informative and normative, and analyzes how diffusion flows through the networks. The informative effect is focused on the speed of diffusion is scale-free networks while the normative effect is focused on the speed of diffusion in highly clustered networks. The informative effect is measured by a consumer’s desire to gain more information. The normative effect is measured by one’s desire to keep up with social norms and imitate what others are doing. The study found that the most important factor for diffusion is the non-adopters that facilitate information to other non-adopters from the original adopter.
This relates to networks and the diffusion throughout networks that we studied in class. If we consider real-world interactions between people, when we recommend products to each other, that’s a form of word-of-mouth advertising. Most people are not the initial adopters of products, but as more and more people begin to use a product and recommend it to their friends, new innovative products find their way to everyday people. In class, we spoke about the threshold and if we translate this to the real world, we can view each person as a node having their own threshold. Each person receives payoffs based on strong links to people they are close to and weak links with other people they interact with that also use that product. The more people we hear talking about this product, the more likely we are to look into it (informative effect). On the other hand, if a majority of people are using a product, it becomes a norm and we feel obligated to join in (normative effect). Through word-of-mouth advertising, we help products gain the users they have. Without this mass of people spreading products through their own networks, diffusion would not be possible.