The diffusion of innovations represents a fruitful topic of research in innovation studies. Different scholars have proposed alternative models explaining the timing of adoption: epidemic models, probit (rank) models, stock models, order models. Empirically, few papers have compared these different models. The aim of this paper is to study empirically the factors affecting the timing of adoption of a consumer technology. We test alternative models in relation to the diffusion of portable Digital Audio Players (DAPs). Our main source of data is a survey of 1562 potential adopters from 8 European countries and Japan. Our preliminary results highlight that, first, in line with the rank models, individual characteristics affect the speed of adoption; second, consistently with stock models, the adoption is subject to strong network effects; third, as in epidemic models, the time of adoption depends on the diffusion of information on the innovation.