With the proliferation of Enterprise Social Networks (ESN), the measurement of ESN activity becomes increasingly relevant. ESN analytics aims to develop metrics and models to measure and classify ESN user activity to support organisational goals and outcomes. In this paper we focus on a neglected area of ESN analytics, metrics for understanding activity in ESN groups. We engage in explorative research, utilising actual ESN meta data describing activity in 350 groups across three organisations, to identify a set of metrics that divides the group sample into distinct user types. By employing clustering techniques we derive a set of four group types: broadcast streams, information forums, communities of practice and project teams. For ESN analytics research we contribute a set of metrics and group types. For practice we envision a method that enables group managers to compare aspirations for their groups to embody a certain group type, with actual activity patterns.