As it becomes apparent that users are an important source in innovation in society and in organizations, scholars are realizing that user-directed innovation policy might contribute to improving social welfare. How such policy might be designed, however, is uncertain, as are the costs and benefits of such policies. It is also not clear whether there is a problem for user-directed policy to solve, or what that problem is. As a first empirical step to answering these questions, we report the results of providing hospital clinicians with access to ‘makerspaces’, i.e. staffed facilities with prototyping tools and the expertise in using them. Findings suggest that almost all innovations developed in the makerspaces are user innovations; that the potential returns from the innovations developed in the makerspaces’ first year of operation are more than tenfold the required investment; and that most of the innovations would not have been developed without access to makerspaces. Due to lack of diffusion, only a limited share of potential returns is realized. This suggests not only that there are problems of non-development and under-development that policy can solve and that doing so supports social welfare. It also suggests makerspaces as an effective form of user-supporting innovation policy.