Danah Boyd’s take on why MySpace worked. The short answer is Subcultural Capital. The longer answer is in an essay on Boyd’s site that I’m excerpting below. Initially the essay sets up a history of Friendster and MySpace which is a good read in itself but moves into reasons for success later.
It is not about technological perfection.
Portability of identity doesn’t matter. Easy-to-use interfaces don’t matter. Visual coherence doesn’t matter. Simple navigation doesn’t matter. Bugs don’t matter. Fancy new technologies don’t matter. Simple personalization doesn’t matter.
Before you scream “but it does to me!” let me acknowledge that you’re right. It does matter to you. The question is whether it matters to the masses. And it doesn’t. Especially for teens.
…
Friendster focused on simple and narrow, giving users very limited options and cracking down on all hacks. For a long time, they took away features rather than adding them. They worked to mainstream-ify, to be equally generic to all users. MySpace added features all the time, making it a game to see what had changed, to find new ways of navigating the site. Hacking the site became a cultural phenomenon with websites being dedicated to hacking techniques (brought to you by fellow cultural participants not O’Reilly). MySpace let users define the culture.
It’s worth noting I found a link to Boyd’s essay and an interesting read about podcasting stats on the Powerhouse Museum’s Fresh + New blog.

Eric
Thanks for the referencing/linking.
We tried a social networking style project here at the Powerhouse Museum way back in 2001 called Soundbyte. Looking back it was a bit like YouTube for high schools – schools could upload class made video to share/stream. It worked as a way of promoting and drawing attendence to our media labs where we teach digital video, audio etc. We closed the site down in 2004 because we weren’t able to offer the kind of networking that users needed/wanted because it was run by a museum bound by policy concerns – we even required teachers vouch for Copyright clearances in the uploaded works rather than letting students upload themselves. Nevertheless the site was pretty popular and did result in marked increases in visitation to our media labs – which was the intention. Some of this real visitation was from people wanting to improve their skills in making content to upload, others just became aware of our facilities through the site.
I think if we did it again today in order to be successful we’d need to be a lot more open in terms of access and trusting of users which is something that is difficult to do within our kind of rightfully risk-averse organisations. I wish that Creative Commons had been around at the time we started developing it – their ccMixter project would have been ideal for what were trying to achieve.
On the issue of museum podcasting I’ve just posted some figures from OSC on their podcasts that may be of interest to your team and readers.