This post also appeared on Communitize, the Favela Fabric blog.
Remember those days when we still lived in caves? When we used to go out hunting or berry picking with our tight knit little community of cavemen. Those were the days. Did you ever wonder how many people these groups consisted of? We did. Why? Because we like to talk about communities. That’s why we looked in to the so called Dunbar Number, a number that predicts group sizes for primates named after Robin Dunbar, anthropologist and evolutionary psychologist. Dunbar was able to determine the maximum size of a group in which any individual can sustain stable relations with all other individuals in that group by looking at cognitive capacity. The bigger the brain, the bigger the group can be without becoming unstable.
Due to our ability to speak, instead of just waving our arms around like a chimpanzee, we can sustain stable relations with up to 150 people. A larger group and we will have to spend the most of our time on social interaction instead of hunting, gathering or watching Friends, thereby greatly diminishing our chances of survival. The question of course is if and how this applies to online communities. Christopher Allen, social software advisor, has written a few very interesting articles on this topic.The key point here is that we do not invest all our time in an online community, and our energy in such a community is not focused on mere survival. So is there another way of looking at the Dunbar number in relation to online communities?
Allen looks at the functioning of groups in varying sizes from 2 (business partner or spouse) to 150. He notes that 7 seems to be a number that functions very well for committees or working groups, whereas 13 seems to be the ‘Judas’ number where close groups start to disintegrate. 40-70 on the other hand seems to be ideal for small business or army units. From 100 upward we have to realise that the community and sustaining stable relations within it will take up all of our time and thus becomes exclusive. When groups become larger than the Dunbar number we have to find ways to restructure those groups, by installing a middle-management for example or using military like hierarchy.
And now back to our online communities. There are a few take-aways from these observations. First of all, it shows the importance of social grooming for communities. Creating stable and enduring relations requires social interaction and not solely subject-matter talk. That’s why, at online fora, we often see a section that allows for gossip or chit-chat. Of course we don’t always need a stable community, sometimes a one time only interaction will do, but you will need to put a disproportionate amount of energy into mobilising and activating people. Events or initiation rites work well in this sense, because they create a shared history. Something that members can refer to and that binds them.
Secondly, there’s something to learn about the size of communities. We’ve mostly been talking about active groups, but the fact of the matter is that online communities often vary widely in activity. In this sense we can fall back to variations on power laws like the Pareto Principle, that 20% of the members are responsible for 80% of the activity, or the 1% rule (or 90-9-1 rule) stating that one percent of the members will create content, nine percent reacts on that content and ninety percent just reads. These numbers can vary depending on the sort of community, but not that much. The conclusion here is that to get a community going with a small working group of about seven power users (the one percent), you will need about 700 members. In this way we can use the Dunbar number to understand that those first couple of hundred members of your online community are vital to its existence and to keep it self-sustainable. Less than that and you will need to continue putting your energy and resources into mobilising and activating new members.
Of course there are lots of exceptions to these rules depending mostly on the intensity with which one participates in the community related subject. You can imagine work related communities or employee-engagement communities having higher participation rates, but even there participation varies widely depending on subject, time and commitment. The Dunbar number then gives us a means to look at the amount of energy we need to put into mobilising new members, activating existing members or in general the scale of a community. We’re interested in hearing your thoughts on community sizes and how you crunch those numbers!