Dave Neary and I co-presented a session about metrics at OSCON on Wednesday based on what we have learned so far from doing the MeeGo metrics.
Description
Every community manager knows that community metrics are important, but how do you come up with a plan and figure out what you want to measure? Most community managers have their own set of hacky scripts for extracting data from various sources after they decide what metrics to track. There is no standardized Community Software Dashboard you can use to generate near-real-time stats on your community growth.
Like most open source projects, we have diverse community infrastructure for MeeGo, including Mailman, Drupal, Mediawiki, IRC, git, OpenSuse Build Service, Transifex and vBulletin. We wanted to unify these sources together, extract meaningful statistics from the data we had available to us, and present it to the user in a way that made it easy to see if the community was developing nicely or not.
Building on the work of Pentaho, Talend, MLStats, gitdm and a host of others, we built a generic and open source community dashboard for the MeeGo project, and integrated it into the website. The project was run in the open at on the MeeGo wiki and all products of the project are available for reuse.
This presentation covered the various metrics we wanted to measure, how we extracted the data from a diverse set of services to do it, and more importantly, how you can do it too.
The