Large and impressive data collection efforts often fail to make their data useful for answering policy questions. In this paper, we argue that this is due to a systematic gap between the ways measurement engineers think about their data, and how other disciplines typically make use of data. We recap our own efforts to use the data generated by a number of such projects to address questions of Internet and telecommunication policy, and based on our experience, propose five points for engineers to consider when building measurement systems to reduce the gap. Ignoring the gap means that fewer researchers use the data and significantly lowers a project’s impact on policy debates and outcomes.