New Bass Model Utilizing Full Social Network Data (Big Data)Speaker : Tae-Hyung Pyo
School of Business State University of New York at New Paltz
The Bass Model has been used extensively and globally to forecast first purchases of new products. Most models for the diffusion of innovation are deeply rooted in the work of Bass(1969). Potential customers may be connected to one another in some sort of network. Prior research has shown that the structure of a network affects adaption patterns. The focus of this study is to explore how network structure affects the estimation of the Bass Model parameters, and how to incorporate network information into the Bass Model.
First, I prove that the Bass Model assumes all potential customers as linked to all other customers. Through simulations on individual adoptions and connections among individuals, I show that the estimate of social influence (q) in the Bass Model is biased downward. Also, the estimate of the external influence (p) is biased upward due to the egative relationship between the p and q estimates. I relax the assumption of the fully connected network in the Bass Model by proposing a Network-Based Bass Model (NBB), which incorporates the network information into the traditional Bass Model. In the simulation study, parameter estimates from the NBB Model are, regardless of the network structure, accurate and provide superior fits compared to the original Bass Model. I further test the NBB Model on large-scale data from an online music social network. The empirical analysis confirms the findings in the simulation study: for all products, the estimates for the social influence parameter (q) are lower for the original Bass Model.
Audience : Faculty, Graduate student
Department : School of Business Administration
Staff : Hye-jin Kim
Contact : 217-3666