Cuil - New Search Engine on the Block

by Vanessa on July 28, 2008

Google has a new competitor. Search engine Cuil, pronounced “cool” (the word ‘cuil’ is a Gaelic word meaning knowledge), officially launched today. And while Google can definitely use some competition, I don’t expect Cuil (www.cuil.com) to revolutionize the search experience.

One thing it has going for it is that Cuil was started by former Googlers, who presumably learned a thing or two while with the search behemoth. But it has some immediate differences. For one thing, Cuil can’t use PageRank to organize the results, as Google obviously has that technology squarely in their clutches and they’re not letting go. So Cuil has to look for different criteria with which to assess the relevance of a page. Cuil actually looks at the content of a page, and ignores the popularity measures that Google considers most relevant, such as link analysis.

First, Cuil finds pages with the keywords you searched for, then analyzes that page’s further content and “its concepts, their inter-relationships and the page’s coherency.” The results are returned not in a list but in a 2- or 3- column format, which is visually a more interesting way to view.

Another nice thing is that Cuil’s home page is black, with no ads or other distracting clutter, and there are no sponsored results at this time, so the user experience is bound to be a little cleaner.

Cuil also offers suggestions for refining results. Cuil claims to have a searchable database that’s 3 times the size of Google’s, but really that doesn’t matter much. The size of the index isn’t the most important thing to a user; if the engine doesn’t return the results the user is expecting, they just won’t use it again, and it won’t matter how many pages Cuil has to pull from.

Another thing that makes Cuil a little bit cooler than other search engines is that they promise not to collect or save user data, so your search history is always private, unlike the big Google.

Still, there’s something missing. I searched for “blackberry,” thinking that would be a good search term to test because of the different meanings of the word. All of the results were about the hand-held device, but to the right of the page was an “explore by category” button with the option “Hybrid rubus,” in case you were wanting info about the fruit. Under that category was three options: boysenberry, loganberry, and olallieberry, all of which are varieties of blackberries. However, clicking on those subcategories revealed that Cuil didn’t have any information about them. Cuil suggested that I entered a typo, that my search term was rare, or that I had entered too many search terms. If Cuil is going to come anywhere near to being the Google-killer they claim to be, they’re going to have to beef up their relevancy algorithm, and quick.

Previous post:

Next post: