Purpose

This 7-year longitudinal study identifies factors influencing the growth of healthcare Virtual Communities of Practices (VCoPs) using metrics from social-network and semantic analysis. Studying online communication along the three dimensions of social interactions (connectivity, interactivity and language use) we aim to provide VCoPs managers with valuable insights to improve the success of their communities.
Design/methodology/approach

Communications over a period of 7 years (April 2008 to April 2015), and between 14,000 members of 16 different healthcare VCoPs coexisting on the same web-platform, were analyzed. Multilevel regression models were used to reveal the main determinants of community growth over time. Independent variables were derived from social network and semantic analysis measures.
Findings

Results show that structural and content-based variables predict the growth of the community. Progressively more people will join a community if: its structure is more centralized, leaders are more dynamic (they rotate more), and the language used in the posts is less complex.
Research limitations/implications

The available dataset included one web platform and a limited number of control variables. In order to consolidate the findings of the present study, the experiment should be replicated on other healthcare VCoPs.
Originality/value

The study provides useful recommendations for setting up and nurturing the growth of professional communities, considering at the same time the structure of the interaction patterns among community members, the dynamic evolution of these interactions and the use of language. New analytical tools are presented, together with the use of innovative interaction metrics which can significantly influence community growth, such as rotating leadership.