Lists. Endless lists. The latest curse of the web are those endless swimming pools of customer data - most popular, most active, most tagged or downloaded. Personally I hate them. They tell me nothing, other than other people’s aggregated bad taste. Worse – they miss one of the Internet’s most subtle and powerful features – the discovery power of networks.
Networks are amazing things. If the eighties and nineties had Moore’s Law framed on the wall, our decade could as easily enshrine the golden rule of network economics. According to Metcalfe's Law, the value of a telecommunications network is proportional to the square of the number of users of the system. In other words, every one benefits disproportionately by the addition of new user. But just what does benefit mean?
Network theorists talk a lot about the power of connections. An extra person in a phone or fax network, is another person that everyone can call. However in my view, the real value is not just that extra person – but their actual address book – the memory trace of all their prior interactions. If that sounds too abstract for you – ask a teenager.
One of the reasons that MySpace is growing so fast is that it is absurdly easy to not only invite people you know, but to deep dive through other people’s networks and add them to yours. Like attracts like. If you are into Goth Industrial Rock, finding someone with similar tastes to yours is like tapping into a vein of pure gold. You can be sure that everyone on their friends list has similar tastes, as will their friends and so on. Before you know it, you will have a hundred friends you have never met.
Compare that experience to Amazon’s personalisation service or Apple’s iTunes. Both services use reasonably sophisticated filtering tools to make recommendations based on other users with similar purchasing patterns, or lists of most purchased items. Despite the slick algorithms, they are tools of limited utility. Even if Amazon doesn’t peg you for gay because you buy your wife cooking books and opera CDs, the mass aggregation of consumer data has the banal effect of turning taste into linear function. There are simply not enough good surprises.
What would be really interesting is if online retailers allowed you to dig through clusters of real customers who bought a particular book or song, and then wade through their actual transaction history looking for clues. The data is all there. The value is not in averaging it, but making the underlying networks more transparent.
Navigating networks is not an entirely new form of behaviour. You might remember the phrase that people rarely use anymore - web surfing. More appropriate to a time when the best of the web could be surmised by the Netscape Cool List, it is also a good analogy for the way that people browse through other people’s social connections and data trails.
Network surfing is the opposite of search, even though both use similar techniques for determining relevance. Search engines spider link networks looking for patterns which they then aggregate into result sets. Network surfers act like spiders themselves, but not with a view to finding a specific answer. If anything, you skim for a vibe, an individual trend, or a cluster of users which they can then mine for the information they want. It is inductive as opposed to deductive thinking. Friar William would be proud.
There are lots of examples of network surfing in action – other than just stalking hotties on MySpace. If you want to become an instant expert on the Chinese internet market in under an hour – using Google isn’t the answer. Your best bet is to find a couple of highly plugged in Chinese blog writers, and surf through their network of sites they read, link to, the people that comment on their site, and their blogs and so on.
Similarly - if you want to figure out the hottest underground music in New York at the moment, typing “hottest music New York” into a search engine or looking at the most downloaded pop song list in iTunes is not going to help you much. Try instead finding a cohort of creatives on MySpace and surf their friends lists, looking at which bands are most linked to and discussed. Or log onto Last.fm, a music based social network. Last.fm plugs into your music player, and by tracking what people are actually listening to, lets you trawl through real data rather than taste guesstimates.
As seductive as mass statistics are – the fact remains - aggregation is dumbing down. The insight is in the details. Using Google doesn’t make you smarter. Using the people that do, does.
