Bayesian and Conjoint analysis on Forecasting the Home-Networking Market of South Korea in Competitive Environments

In this paper, we develop prelaunch forecasting model reflecting competition structure among various technologies and this model overcomes limitations of the Bass-type diffusion model which relies upon historical sales data. This paper applies our model to the HomeNetworking(H-N) market where there exist various alternative technologies in South Korea. First of all, we develop a Bayesian model based on prior information derived by expert judgments. In this model, new sales data are incorporated into the timely update of forecast. The result shows excellent fitting ability between actual values and estimated values after Bayesian updating. In addition, to describe competitive environment among the various technologies, we use a conjoint analysis based on our survey data of technology choosers. By incorporating dynamic indicators of H-N technology into our conjoint model, we also forecast changes in preferences of technology choosers. The forecast shows that the market share of Wireless LAN is the largest among the H-N technologies at any time. Finally, based on the simulation experiments, we also identify important factors that affect the demands of H-N technologies. The results clarify that cost reduction, standardization, consumers’ preference, and the government policy affect significantly diffusion of H-N technologies.