Online music network
http://fortune.com/2013/07/18/the-forgotten-music-network-actually-boosting-sales/
This article is not written recently, but I find some interesting concepts in it. Basically, this article introduces a scrobbler called Last.fm. After being downloaded, it will keep track of when and what the user listens to. It also forms a social network for its users, so that they can find people with similar taste or even share music list together. This scrobbler, surprisingly, correlates more highly to music sales than many social medias do, like Facebook and Twitter.
The CEO explains that people using the app are all music enthusiasts, so that they really care about the music and are more likely to buy the album. However, don’t Rihanna’s fan group on Facebook and followers on Twitter care about her music? Why don’t they buy the album then? I believe we can use tie theory and imaginary networks to solve this problem.
The fans network of a musician on Facebook or Twitter would look like an ego network, where there’s a node in the middle (the musician him/herself) and many other nodes connecting to it. In this situation, edges in this network should mostly be weak ties; there are no chances for them to know each other, and there probably won’t be many interactions (namely, no edges) in between the ego’s fans. On the other hand, if we create a fans network on Last.fm, regardless of its type, there will be more strong ties with in (in other words, if we make it an ego network, it will have higher clustering coefficient). Why? Here is the reason: even without the musician him/herself promoting, the built-in function collects data and match people with similar tastes, and through interactions like sharing music lists, they are more likely to develop strong ties, and thus motivate each other to purchase the music.
Since we are also talking about positive/negative relationship in class, I also find it interesting of how online music network may be relevant to this topic. This does not mean that there are both fans and haters, but the fact that even when there is an edge between the musician and a fan, we don’t necessarily know the property of this relation. Currently, our model of this network is unable to predict how much a fan can do for the musician. Does the fan simply want to follow, to listen to the music online, to buy the album, or even persuade others to buy it? I guess this is a difficulty of modeling social networks.