Comparison of diffusion models for forecasting the growth of broadband markets in Thailand

The aim of this paper is to investigate the most accurate S-curve model, the Logistic, Gompertz, and Bass models, in forecasting the diffusion of telecommunication markets Thailand. The analyses apply the data of mobile telecommunication market and fixed-broadband market separately without the interaction between both services. The originality of this study is at the diffusion path segmentation intervened by technological change that accelerates or decelerates each market. Parameters of each model are estimated by nonlinear model estimation methodology. By applying those parameters, the accuracy of each model can be identified compared to the actual data. Following the evaluation of the goodness-of-fit and forecasting ability, the Gompertz model shows the best performance in forecasting the diffusion of mobile telecommunication and fixed broadband markets. With the more suitable forecasting model to the markets, the ultimate total number of users in the future period could be more accurately predicted.