In this month’s Fortune Magazine, there is a great article on “Finding the next Google,” in which they talk about the next big thing on the web. They believe that Google rode the wave of finding to incredible profitability, realizing that people would use the web to find specific items or bits of information. The future, according to the article, is in discovery,that is showing you things based on your current likes and dislikes that you are not currently aware of but will probably be interested in.
The most obvious examples of success in the discovery space are sites like last.fm and pandora, who are both building extremely popular communities
around the concept of music recommendations (discovery). They know that if person A loves Jack Johnson and Dave Matthews, and person B loves Dave
Matthews then they will probably like Jack Johnson, and they’ll then play a song of Jack Johnson for person B. These optimized music introductions are incredibly effective at pushing your music limits very quickly because all the new songs you are exposed to have been carefully screened…by others.
This is also clear on Amazon.com, where they have watched the buying habits of others and have your buying history handy, so they can easily create targeted, automated product recommendations to you on the homepage instantly. This creates a far more effective landing pages, than just a bunch of categories or instant products. This is about personalization and automation.
I believe, as Fortune mentions, the next wave of the web lies in the ability to assist in targeted discovery. People want to explore the endless supply out there, but have no idea where to start. It’s extremely complicated to ever think of recommendation engines knowing if you like BMW’s then you’ll like Starbucks and Kenny G, but with enough connections (ie connections to your existing personality online…myspace, email, amazon account, last.fm account, ebay account) and enough computing power, it’s possible to start building comprehensive personal profiles, which in turn would allow for more and more relevant recommendations. Imagine taking it a step further and offering a last.fm like scrobbler, that sits in the background of your internet browsing quietly building an anonymous (anonymous in the sense that NO ONE ever would have access to it with your name on it…it would only be viewed by machines for the purpose of recommending things to you) profile that could offer you the perfect vacation designed just for you, the ski jacket you’d love, and a news site right up your alley. I’m in a situation where people are asking what I want for Christmas, and I really have no idea what’s even really out there. If had a stream of recommendations, it would be easy for me and everyone I want to share those recommendations with, to know instantly.
Another way these profiles could be built (and maybe this is why google is offering discounts to people using google checkout this holiday…to build a user base and profiles) would be to have access to, and analyze your real life spending. I would imagine that most purchases over $30 are made with a credit or debit card, which of course means they are entirely electronic and can therefore easily be analyzed for patterns, and used to build a preference profile around you. What if google check out was an easy, back end way of doing this? If they got enough people buying across all sorts of sites, they’d be able to start building recommendations as well. Or even more possible, would be someone partnering with a credit bureau to build personalized shopping recommendations based on current credit and past buying history…just rambling here.
It’s a scary tight rope to walk between big brother and making sense of all that is out there, but I think we’ll get there in a way that ultimately proves to be a major win/win for companies and consumers alike.
Another company moving into this space is Aggregate Knowledge, see them here.
Life Putting 