09 March 2006

The web 2.0 love affair continues...

It's only a matter of time before Web 2.0 becomes sick of my fawning and tells me to leave it alone, resulting in me hiding in its garden at night and stealing underwear from its washing line. But until then I will continue the puppy-dog eyes thing.

Today's great looking bit of kit has the potential to kill MySpace, hurrah! Well, maybe not, but it's an interesting rival to it. This New Scientist story tells of an automated tool for web-page creation which uses key words parsed from user-suggested feeds, likes etc to generated a page containing other relevant news.

It's along the lines of things like Last.fm, doing something which the internet is becoming really good at - opening people's eyes to something new. The concept of going to sites and following lots of links to vaguely relevant things and eventually ending up somewhere cool is hardly new, but doing it automatically is great. Last.fm's power lies not in advertising to other people what you like (which in itself has interesting applications) but in seeing what else you might like. As an example, I keep getting reccomended Death Cab for Cutie and will sooner or later have to buy one of their CDs.

Amazon have been doing something similar for ages, and a "you might also like" link on e-commerce sites is now pretty standard. But they're easily fooled; if I buy one book as a present for my brother suddenly my results are skewed into things I have no interest in. If (as I do) you use e-commerce for presents, having a firm idea of what someone wants, and buy things for yourself in real shops, then it's next to useless.

The Last.fm concept, and this new idea, are bound to give far more relevant results because there's just enough user interaction and choice involved for it not to be 100% automatic. Automatic is good but it's easily fooled. And having that feeling of having direct control over results - and that results aren't skewed - lends it more credibility and makes a sale more likely. I'll trust Last.fm's reccommendations because I know it's based around what I'm listening to now. I won't trust Amazon's.

I'm looking forward to seeing this go live - I wonder how I can get an invitation?

No comments: