What’s in a Group? Identification of group types for Enterprise Social Network Analytics using SWOOP data

Abstract

We report on research, carried out in collaboration with SWOOP Analytics, to identify metrics that allow distinguishing groups in Enterprise Social Networks (ESN) according to their activity patterns. The emerging field of ESN Analytics has made inroads into providing metrics and models to measure 1) the health and structural properties of enterprise social networks, as well as 2) the activity pattern and distinct behavioural roles of individual users. What is lacking so far is ESN Analytics at the group level. Yet, groups play an important role in ESNs for organising communication and collabo-ration activity. In this study we carry out explorative research employing cluster analysis to identify metrics that best distinguish a sample of 350 ESN groups from three organisations into distinct types. We identify three metrics as most useful: 1) the Gini coefficient, measuring (un)evenness of user par-ticipation, 2) density, measuring the extent to which users interact with each other, and 3) reciprocity, measuring the response rate to messages within the group. The resulting typology of four groups, broadcast streams, information forums, communities of practice and project teams, will be useful for network managers and group leaders to check how well their group is tracking against intended group activity pattern.