billk2

The new search engine hits the streets, but misses Surbiton.

The big search engine news this week is Cuil (pronounced Cool). Apparently Google employees going it alone. I'm not sure where their expertise lay, but PR seems to be a strong point, they've got an inordinate amount of publicity considering.

So how does it fair? Well the big difference is that Cuil indexes a hell of a lot of pages. Billions. Much more than Google. In practice however, the results are questionable at best. Disregarding the example of the university professor's results page that's currently littered with porn images, personal experience is also less than on par.

One search term I'm familiar is my home town name, Surbiton. I've monitored these results for many years, across many search engines because I run a website dedicated to Surbiton. I'm pleased to say my site takes the number one position on Google and most of the other top engines. Cuil however demotes my site to page four and for some reason thinks a slalom skier is an appropriate image to associate it with. The page Cuil decides is the most relevant page for anyone searching Surbiton is to a page inviting people to an Hellenic dance group, last updated in 2004.

So what? We probably all agree that competition is a good thing, and that however un-evil Google claims to be they have a stranglehold on the market. But as far as I'm concerned I can't trust Cuil's results and they've lost me on their first try. Unless Cuil can get the same publicity once they've fixed their problems I'm unlikely ever to return. The web is famous for it's ability to push out iterative improvements, but for some applications you absolutely have to get it right first time.