Achieving Critical Mass in Social Networks
Link to paper: Social Network Critical Mass
In this paper, author Chris Geddes discusses the notion of critical mass in a network, and the applications of this concept to facilitating the adoption of social networks. It explains that a fundamental difficulty in increasing adoption of a social network is a sort of chicken and egg problem: people need to visit the site to give it content, but content is needed for people to visit the site. The article cites that a result of this is that generally the rate of adoption of a social network in a community is very low until community penetration reaches about 15%. Here, community penetration is the ratio of the number of people in the community on the network to the total number of people in the community. Once that 15% saturation is reached, it’s observed that the rate of adoption rapidly accelerates until the community is fully saturated. As a result of this, it’s often easiest to start off by releasing a new social network in a small community, then repeatedly doubling the size of the community which has access to the site so that after each doubling, the total saturation percentage is greater than 15%. The article takes the adoption of Facebook as an example of this. When Facebook was first only made available to Harvard students, a relatively small community which could more easily be saturated. After saturation was reached, Facebook was opened up to a second university, and so on until a large number of universities were saturated. Only then, was Facebook opened up to use by the general public. In this way, Facebook’s community saturation rate never fell below 15%, and was thus able to maintain a high rate of adoption.
An interesting point brought up by the article is that in parallel with the increased rate of adoption at 15% saturation, something else interesting happens: the likelihood that the social network is fully connected jumps from ~0% to ~100%. This ties in to the concept we learned in class that most networks (for example Facebook) are composed of a single giant component or, in other words, are totally connected. This meshes very well with intuition: if a social network is totally connected, there’s a sense in which it’s more useful to each node (person) in it because there’s a wider group of other nodes (people) he/she can communicate with, which is the whole point of a social network. So, once this 15% threshold is met, the graph becomes fully connected, and thus more appealing to potential new users, accelerating the rate of adoption. As was mentioned above, Facebook is a great example of this. In class, we learned that over 99.9% of Facebook users are contained within the giant component of the network graph, which corroborates the idea that past a certain saturation rate, social networks with very high probability will become (almost) totally connected.