Many technologies require local experimentation before individuals can make successful adoption and usage decisions. Since information is a public good, technology adoption subsidies are an important component of development policy. We ask whether it is possible to enhance technology adoption schemes by targeting the most appropriate experimenters. Building on Chassang et al. (2012), we design and implement a set of field experiments that let us study the building blocks needed for successful targeting: Are people heterogeneous in their ability as experimenters? Is this information known, and by whom? Can it be elicited and how? Can it be used to enhance the effectiveness of subsidies? Joint with Pascaline Dupas, Catlan Reardon and Erik Snowberg.