Identity formation is a fascinating thing in general, and in the burgeoning online social networking sites it takeson an additional angle. Danah Boyd talks a bit about it here for reference.
On most social networking sites, you get standard fields of input for defining your identity in addition to name, etc: 1. Favorite movies, TV, books 2. something about your self, 3. where you went to/are in school, 4. job/income. Typical ways that you might think of defining a person. But I am not sure that really means much – and Danah also suggests that it’s the network that actually means the most (particularly for younger and older folks).
In the music social space, I think it means even less. Think of a band scene. While such a scene will often be populated by a similar set of demographics, there will be some scenes that cut across demographics. Now extract the music from that scene and I think the music will cut across even more demographics. So now, in that case, in terms of music preference, what does identity mean?
You might assume that certain preference groups will spring up – clusters (eg: social network) of people around a specific topic (artist in this case), but I don’t think that’s the case. Or you might think these cluster would evolve around a particular music style. That might be closer, but not quite there. Think of the groups on MySpace around an artist – although they might have a large membership, first they are transient and second the group helps define the user.
But attempting to cluster people artificially around an artist, or genre, or year, or label just won’t scale. It might work a bit for a while or in limited functions, but the “fan club” only goes so far around a specified target.
So, I suspect there is going to be one consistent driver for social formation around music preferences. And it’s not artist, or genre (where even top 40 is a genre). My question is this though: Is the driver a music related item (ie: tempo, tone, etc) or a social item (how it makes me feel, etc), or in the social ties (I like it since it came from her/him)?
Once you define music preference and identity in terms of the social network, things change quite dramatically in thinking about music delivery. Why on earth would I assume that you would like something because it’s popular? Or because you have a lot ofbluegrass music on your computer? Or because you have listened to a lot of Arctic Monkeys? “This artist sounds like that artist, so you might like it.” It just doesn’t mean a whole lot.
Don’t get me wrong, it does work a bit, but I think the social stamp on music preference is the real killer. But not just because someone else is listening to it right now.
It’s because someone recommended it you. The strength of the social tie is critical and the recommendation demonstrates it. And when your identity is built on your network, your ties are everything. So – more successful recommendations mean stronger ties and and stronger ties clarify your identity.
And once that is clarified, we can now get down to the business of accurate recommendations and predictions on song growth. But I am not going to tell you how we do that.