We introduce an international, adaptive diffusion model that can be used to forecast the cross-national diffusion of an innovation at early stages of the diffusion curve. We model the mutual influence between the diffusion processes in the different social systems (countries) by mixing behaviour. Furthermore, we apply the matching procedure as proposed by Dekimpe, Parker and Sarvary (1998). This international diffusion model is adaptively estimated using an augmented Kalman Filter with Continuous States and Discrete observations, developed by Xie, Song, Sirbu and Wang (1997). This is the first application of this procedure in an international context. We empirically applied this method to the diffusion of Internet access at home, and mobile telephony among households in the 15 countries of the European Union. The results show that our international, adaptive model performs well and is by far superior when compared to the classical method of estimating diffusion models for each country separately.

bayesian estimation, cross-country diffusion, forecasting, international marketing
Statistical Decision Theory; Operations Research (jel C44), Business Administration and Business Economics; Marketing; Accounting (jel M), Marketing (jel M31), Innovation and Invention: Processes and Incentives (jel O31), Comparative Studies of Countries (jel O57)
ERIM Report Series Research in Management
Erasmus Research Institute of Management

van Everdingen, Y.M, & Aghina, W.B. (2003). Forecasting the international diffusion of innovations: An adaptive estimation approach (No. ERS-2003-073-MKT). ERIM Report Series Research in Management. Retrieved from http://hdl.handle.net/1765/1093