It is well known that there are many stochastic factors in the diffusion process of technological innovations. To overcome the shortcomings of the S-shaped diffusion model, it is necessary to build more practical models from the stochastic point of view. On the basis of an analysis of firms’ stochastic adoption behavior, this paper builds a stochastic model for a diffusion system of substitutable innovations and performs a computer simulation. Furthermore, in this study, an attempt has been made to explore how to define the parameters of stochastic diffusion models. Thus the approaches to control and optimize the diffusion process become clearer.