This article is about progress in social measurement. I argue that, to improve measurement practices, we must supplement the technical agenda and the conceptual agenda with a politico-administrative agenda. The case study is a reform programme of calculating Gross Domestic Product (GDP) - abandoning the zero-hypothesis for government activity - the implementation of which is lagging behind. The article seeks answers for the implementation problems in the reform’s statistical impact on GDP growth, the institutionalization of GDP, and the potential for gaming. The article concludes with a discussion on how measurement agendas can be better combined in order to make progress in social measurement.