Important ideas, practices and artifacts often fail to reach their target population efficiently or fail to reach altogether. Surprisingly, most projects aimed to bring technology to underserved communities of the world lack an explicit diffusion strategy and/or lack an implementation strategy that acknowledges the social structure that binds together the members of the targeted community. Without the knowledge of social structures efficient diffusion of technological innovations becomes an unreachable goal. Socioeconomic and behavioral information can be combined with sparse social structure data to derive quantitative estimates of a community’s social dynamics, allowing improved understanding and management of diffusion processes. We found that patterns of advice and use of media provide and effective way to identify the influential members of a community. We set up a large scale experiment in a rural community using our model and tested our proposed method of intervention and found strong evidence of an improved diffusion process which is significantly related to the communities’ network of advice. The adoption of an idea, practice or artifact is heavily influenced by social context, through both conscious and unconscious mechanisms. By targeting social networks, not social classes, age, gender groups or institutions, we can create the basis for the emergence of local organizations and businesses that organically provide the necessary support to achieve effective diffusion of technological innovations. In our experiment, the introduction of a few powerful ideas at the core of the communities’ social networks helped to create a social context where the new innovations created economic and social value. In addition, these new business create a richer social context from which further new innovations are expected to emerge.