We are getting some really great pickup on the Community Health Index thanks in part to Jeremiah Owyang and his blog. One question that surfaced recently: "Is the Community Health Index dependent on size, age or typology (form and function) of the community. Answer: No. The index is robust in this regard. Of the 6 heath factors – those which you use to action decisions, 3 are specifically size independent: Responsiveness, Interaction and Liveliness, the other 3 are size dependent. The Community Health Index itself is normalized in such a way that size, age and typology each are taken into account by looking at the change in these characteristics over time, discounting outliers (like the edge effect of a community launch on registrations, seeding from data imports, and spikes due to campaigns). There is a decay function over 50 weeks that accounts holds onto the memory of recent activity, but historical changes loose relevancy. I like to think about it this way: Just like for an individual, a single workout may indicate an effort towards overall health, but the sustained and steady improvement of the factors is what really counts: Community Health must be managed, in a sustained systematic manner without major fluctuations to achieve maximal health. We see examples of this with registration drives to attract new members. A spike in registration may effect Members and Liveliness for the short term, but these effects are discounted in the calculation of your Index if the effort is a spike. In fact, the volatility caused by such a spike is often a detriment because of the disruptive nature to the overall continuity of the community. With regards to the type of community (support, marketing enthusiast, innovation, etc) we do see differences in the individual health factors. For example, Interaction for a support community typically ranges between a value of 2 and 4 for its health factor, but enthusiast communities Interaction is above 6. What is really interesting is for companies who have the intention of running an enthusiast community and the data tells us that the behaviors displayed by the members suggest otherwise. Thanks and keep the questions rolling. Neil @nbeam
... View more
