First, I firmly believe that successful online organizations can be identified by looking at three primary characteristics:
- Distributed Authoring – humans adding content and the wisdom of crowds
- Strong Subgroups – meaning active committees under 150 people typically
- Transparency – a level playing field must be in place for all with controls
These were first articulated in "Engaging Your Membership: What Are You Doing and What Should You Be Doing?" and the Distributed Authoring bullet was expanded in "The Concept of Distributed Authoring for Membership Associations ““ Getting Your Association to “Virtualization“. And of course everything we program in Tendenci is designed to facilitate these three organizational goals. But at the end of the day it is up to the association to determine the action and policies it will demonstrate.
Some data. Here is one graphical snapshot from a randomly selected Tendenci client in aggregate. I changed the numbers a bit, but in a statistically consistent way so the trends are valid.
I filtered out stuff like editing a profile or registering for an event as those are more data entry in my mind. Authoring means contributing an article for the newsletter or posting an event on the calendar (again this is subjective and my opinion).
I did try fitting a logarithmic and an exponential distribution, but the power curve was the closest match despite the divergence as it approaches zero.
This last graph is a pie chart limited to people who actually added content. Again the data has been changed a bit, but not much, so the trends are consistent with the actual distribution. Note again that there are two or three super users adding most of the content that is read and absorbed by the entire membership.
Specifically the top 5 users are adding almost 85% of the content.
One possible explanation is that someone is functioning in an administrative role (not the security level but the act of functioning administratively) with others emailing articles and society events to post on the site. This is likely in my opinion based on observing interactions and I made no effort to correct the data for author versus typist.
A possible future post or article should probably look at which articles and events are being read the most. Something along the lines of what AttentionTrust is interested in as long as it can be done anonymously for the users (nobody likes big brother, especially me!). Thanks!