Online communities are increasingly being deployed in enterprises to increase productivity, improve employee skills and job satisfaction, share expertise, and enhance innovation. However, attaining these benefits is no simple matter and an effective community leader is a critical success factor. Because of their essential role in fostering community success, online community leaders are a growing population in enterprises, but existing technologies rarely support leaders directly, both because of a lack of clear data about leader needs, and because existing tools are member- rather than leader-centric.
Our aim is to develop and evaluate novel tools that facilitate effective leadership in online communities. Working closely with community leaders, we co-developed Community Insights, a novel tool that provides useful, actionable, and contextualized analytics to help community leaders foster healthy communities. We deployed Community Insights to leaders of 470 communities in one enterprise over 10 months, gathering a detailed understanding of community analytic needs in interviews (with 26 leaders) and from Community Insights log data (with nearly a thousand users). Our work with community leaders highlighted the following:
(1) We need to rethink community metrics for the enterprise. Prior research based on internet communities emphasizes growth in both volume of posts and community size as key community health metrics. In contrast our interview and usage data showed that enterprise community leaders were most concerned to see views of community content as a health indicator. Overall however, leaders doubted that simple metrics alone were particularly useful, as communities have very different goals and are at different phases of their existence. Instead, leaders found more value in metrics linked to actions that helped them facilitate community goals. Our results suggest the need to develop new metrics focused on achieving specific goals within the community, rather than focusing on overall growth measures.
(2) Community leaders want actionable metrics. Community leaders were not just interested in passively understanding raw data; they wanted Community Insights to enable them to determine remedial actions. Leaders highlighted four types of information that led to action:
- Identifying particular sub-groups or members for active engagement, such as asking topic experts to contribute new content, soliciting feedback from people who had left the community, and tailoring content to sub-groups in different geographic regions.
- Identifying community health problems that require facilitation, by presenting deviations from typical activity patterns or baseline “healthy” thresholds.
- Identifying successful examples to emulate (e.g., popular posts or topics), both inside and outside the community.
- Evaluating effects of particular actions or events in the community to aid future planning, e.g., by examining activity trends in the time period surrounding the action or event.
(3) Contextualized analytics are critical to support sensemaking about community data. Prior work neglects to discuss critical challenges in helping leaders make sense of community data. Sensemaking is not just about usability (e.g. readable charts), it is about presenting analytics so leaders can understand their application. To address this, we developed contextualized analytics in Community Insights, data presentations that provided adequate context to support interpretability. For example, if a leader can compare analytics derived for her community with those derived for a set of similar communities, her own analytics will be placed in a meaningful context that helps her make sense of the data. We propose six ways designers can contextualize community analytics: comparing with related communities; comparing with meaningful, general baselines; relating to critical events; relating to leader interventions; helping leaders spot exceptions; and relating to the goals of the community.
For more information about how Community Insights supported useful, actionable, and contextualized analytics to support enterprise online community leaders, see our full paper, Community Insights: Helping Community Leaders Enhance the Value of Enterprise Online Communities.
Tara Matthews, IBM Research – Almaden
Steve Whittaker, UC Santa Cruz
Hernan Badenes, IBM Research – Almaden
Barton Smith, IBM Research – Almaden
Michael Muller, IBM Research – Cambridge
Kate Ehrlich, IBM Research – Cambridge
Michelle Zhou, IBM Research – Almaden
Tessa Lau, Willow Garage