New technologies often enter the market at a competitive disadvantage. While they may seem to promise future advantages such as lower costs, environmental friendliness, or higher performance, initially they may be significantly more expensive than incumbent technologies or face teething problems. The adoption of a new technology may also be affected by consumers’ uncertainty on its performance. In such an environment, information feedbacks are likely to arise and could allow a certain technology to be the dominant one in the market. Drawing on recent percolation models of diffusion that combine the contagion aspect of agents distributed on a network, with the heterogeneity of agent characteristics, we develop a complex-dynamics model of new technology diffusion. By using a multinomial decision mechanism to model each adopter’s choice on a portfolio of new available technologies we show the effect of information feedbacks on market dominance. Using agent-based simulations we explore when a limited subsidy policy, combined with a power-law learning curve for the price as a function of the cumulative number of adopters, can trigger a self-sustained diffusion of a certain technology.