Nat Torkington has a great post on homophily and social software design.
- The Washington Post has a brief article called “Why Everyone You Know Thinks The Same As You“. In short, you hang out with people who are like you, a phenomenon known as homophily. This happens online, and indeed the Internet can lower the costs of finding people like you. But homophily raises the question for social software designers of how much they should encourage homophily and how much they want to mix it up.
He goes on to explore the tradeoff of fulfilling our desire to group with the similar and expose people to the different. I’d suggest considering that the most productive social networks have a dense core and a dynamic periphery, and strong ties don’t come cheap, there is a role for both.
Nat suggests recommendations (if you liked X, try Y) are quick wins, but still within narrow interests, while Mavens-as-algorithms (people who liked X also liked Y) can serve as connectors. Personally I prefer designing in popularity indexes because it inspires productive gaming for participation. I also fear my Gmail thinks I am gay. But while you can automate some aspects of social discovery, they lead to weak ties at best. The answer may be more pivotal…
- Another way to build in serendipity is to have pivotal navigation: tags, top ten lists, and Flickr’s interestingness measure are all ways to break people out of whatever group they’re in and take them to something new. Links are at the heart of this: we’ve all been lost in clicking our way through a drunkard’s walk of the Internet at one point or another. Inspire that in people: build those links and the metadata behind them into your site from the get-go.
Links remove barriers to our abundant desire to share. Not only do we have a desire to be with people we like, we suffer from the problem of believing what we write and create is more valuable than it really is (cough). Some facet of our identity is just waiting to show off to strangers, not just in hopes of finding more people we like, but affirmation from even people we may dislike (deep down George W wants Kim Jong Il to like him, really really like him). We may want to be with people we like, but we have multiple facets of our identity, so we are like different people when we can get away with it. We also affiliate ourselves with people we would like to be, or objects that help express given facets of our identity.
More practically, people and the social incentives that drive them are incredibly diverse, and value accrues to people who bridge social network clusters. So give users tools to bridge divides and create new groups. Support ridiculously easy group forming. While many things drove Myspace into popularity, easy group forming around independent bands drove not only growth overall, but rich serendipity.
I believe people self-correct for homophiliy. Just yesterday in an internal blog, Kirsten Jones noted she realized has been so heads down at work that she was going to subscribe to some new feeds on the outside.
But I have one point for this post, it is that people are better connectors than algorithms and if you give them tools for it, their practices will exhibit emergent properties you couldn’t predict as a solution for serendipity.
I’ll end this post with an underscore, from IWB at IBM:
- Through experience, we have learned the limitations in our ability to leverage IT to automate quintessentially human tasks, even many seemingly simple ones, no matter how fast our computers have become. I think that the key breakthrough that we have needed here is a cultural one – it is perfectly OK to integrate people in the design of our systems. A good, elegant design is one that lets machines do what they do best, and lets people do what they do best.
Tag:
Add to 
Bookmark Murdok:
Ross Mayfield is CEO and co-founder of Socialtext, an emerging provider of Enterprise Social Software that dramatically increases group productivity and develops a group memory.
He also writes Ross Mayfield’s Weblog which focuses on markets, technology and musings.
