(This was accidentally published in "note-to-self" form. Excuse me. It must have seemed more than a little incoherent.)

Anyway,

In a recent paper, The Significance of Social Software, Dana Boyd makes the following point:

"While early social technologies were about finding people with similar interests, the latest round
is far more about connecting to people and watching shared interests emerge
through that."

I do get this. I've been having a great experience with Facebook recently. They are focused on connecting to such an extent I can only image it must be written on the walls above the water coolers, in the header and footer of every document and email, and inserted subliminally in the background office hum, down in their Palo Alto offices. They certainly appear successful. It must be good advice.

How good? Dana also points out the following:

"...those who joined Friendster and assumed that everyone was like them did so
because of the way the site was designed – the structure is inherently egocentric.
This is also where things get tricky because egocentric communities cannot support
that many different contexts. And thus, what you see, is people using multiple sites
to keep contexts separate."

Clearly, defining activities shape the environment in which they are pursued. Though there are "social networking" sites that are successful and focus on the activity, or context, first -- librarything, linkedin, and a lot of dating sites, come immediately to mind. Somehow I don't think LinkedIn would work as well if it looked like a WOW fan site. I think that's the point she's making with context -- that, and the interaction styles individuals pursue within a given context. I, for instance, did not choose to make my LOTRO main character name prominent in my LinkedIn profile.

I wonder where the balance lies. Along those lines, Dana also makes this point:

"The problem is that monetization is hanging on the tip of everyone's tongues again.
To make money, sites have to grow. To grow, they have to expand beyond
comfortable context borders."

It's equally hard for me to imagine a large number of profitable "people-connect-first-and-we'll-see-what-context-turns-up" sites, as it is for me to imagine properties for every possible niche context. I believe that's the balance social experience builders have to address, and keep addressing. The creative in me wants to think you can have your cake and eat it too. I wonder if we could construct an experience that flexibly manages multiple contexts? The optimist thinks the answer might be yes.

I think Dana is asking the more general question here:

"Are there ways to rethink the scaling process to make social software more economically viable without killing the communities in the process?"