Enormous progress is being made in the design of metrics to assess the consequences of social interventions. But four considerations could still usefully be given greater salience in the design of social impact measurement systems. In the first place, it is important to recognize that metrics have consequences, and not all of these consequences are benign. Therefore, we must pay more attention not just to the metrics, but also to the strategies and concepts that underlie them. Secondly, this implies greater sensitivity to whose priorities the metrics should serve, and particularly to the relative weight to place on the priorities of private investors and those who are incentivizing their involvement. Third, care needs to be taken to make sure that the unit of analysis used in gathering the needed evidence is capable of providing real insight into the nonfinancial impacts being achieved. Finally, we should follow Google’s lead and make more room in the design of measurement systems for the perspectives not just of investors and the ventures they support, but also of the beneficiaries of their actions.