Over the past few days, the blogosphere has been, well, blogging, about a new search engine developed by a couple ex-Google employees. Known as “Cuil,” which is apparently a Gaelic word for “Knowledge,” the search engine is attempting to be “the next Google.” (Side note: In case you didn’t know, it’s pronounced “Cool,” not “Quill.”)
I think the site has potential, if they put some serious work into it, but I don’t think they’ll be able to topple Google from their high throne. Google is, and probably will always be, the king of the serch engines. No other search site has become a synonym for “search.” No other search site has ever become a common colloquial verb. (Have you ever heard someone say “I Yahoo-ed it to see what it meant”?) Then you have the Google muscle memory, where you automatically type “Google.com” when you need to search something, the tight integration in Firefox, and all the other services in the Google empire (I’m so used to GMail/Google Apps I can baarely stand “normal” mail clients). And that’s leaving out the fact that I’ve never found a search engine that works as well. Cuil may one day become a major player, but it’s not going to take on Google.
Cuil has an innovative new interface, featuring three columns of search result blocks, which seems to be an effective way to display results, and “related” images are displayed alongside the results. They also have tabs along the top, which are supposed to lead to results in similar topics, and there is also a “drill-down” box to help you refine your query.
The site passes the “Harry Potter Test,” a search-engine testing technique I unconciously started using whenever I demo a new search site. Essentially, the first result for a search of “Harry Potter” should be author J. K. Rowling’s site, rather than Warner Brothers’ cheesey movie sites, which should come later. The fan site called “The Leaky Cauldron” should be high up in the list, with Mugglenet following. An Amazon page and a Wikipedia article should be in the results somewhere too, for bonus points.
I think the search layout may encourage reading beyond the first two results, a problem that plagues webmasters. I don’t know whether it was an intended effect, or not, but I like it.
Moving on to the backend, Cuil claims to have the largest index on the planet, larger than Google even. Though I doubt there’s any way to verify the fact, it doesn’t seem to have as many of my posts as Google does… (“Don’t block firefox,” which I rank 1st for in Google, doesn’t have mine anywhere!)
Cuil seems to be taking less of a link-centric approach than Google, trying to comprehend the page text in a pseudo-semantic-web type of thing. If they can improve their ranking, they’re on to something.
Some of my complaints about Cuil are:
- It’s not stable. It’s slow at times, often doesn’t display any results, etc. I know they’re working on it, but downtime is not an option if you want to compete with the big G.
- The results could be better. Sure, Cuil passes the Harry Potter test, but you haven’t indexed a depressingly large amount of my content, and a lot of SERPs just don’t give me what I’m looking for.
- Why don’t the images shown next to results make any sense? Search for some public fugures and there will likely be pictures of other people in the results (try “Molly Wood” from Buzz Out Loud, for example).
- What’s with all of the Dandelife.com results cluttering the SERPs?
- The main page. It has a black background. It’s hard on the eyes, and contrary to popular belief, it doesn’t save any electricty (it uses more in some cases).
Cuil definitely has room for improvement, but I think it could become a major search player if they play their cards right, though not as big as Google.