Cuil - Google Killer or Big G Roadkill?
August 22nd, 2008 by KrunalA couple weeks ago a new internet search engine named Cuil.com burst onto the internet scene fueled by media hype and the celebrated resume of it’s co-founders, husband and wife team Tom Costello and Anna Patterson, the latter of whom was formerly one of Google’s top engineers, as well as a couple of other ex Google top-level employees Russell Power and Louis Monier.
Although Cuil launched July 28th, I wanted to wait a couple weeks to give them a chance to iron out any bugs before forming an opinion as it’s debut was marred by it’s server crashing due to overload as well as the search engine returning photos from competing sites and attributing them to the wrong site. Naturally this lead to people in the SEO world jumping on the bandwagon to predict it’s early demise, however I wanted to take some time to let Cuil get it’s act together before really evaluating it’s potential… after all, Rome was not built in a day and lest people forget, neither was Google. I know because I was a very early adopter of Google and began using it within months of it’s launch on September 21, 1999 and while better than anything else around at the time, I can assure you Google was not nearly as refined an engine as the Google of 2008.
Ok, so now that I have given this startup time to get it’s act together and refine it’s results, how exactly does it rate in terms of actual performance?
Well, it seems that the photo attribution glitch has been corrected and there have been no more reports of server crashes. As for the returned results, that is a mixed story.
In several tests I have performed on various search terms, Cuil, who claims to have indexed more web pages than any other search engine (over 120 Billion) returns relatively few results for keywords that should have substantially more results. Also, in terms of the relevancy of the SERP’s (Search Engine Results Pages), Cuil seems to be hit or miss, with them returning irrelevant results seemingly as many times as they return relevant ones. This is a difficult thing to quantify as one could enter 1 Million search phrases and still be scratching the surface in terms of the amount of actual searches done monthly on the myriad of search engines around, not to mention for a true quantitative analysis you would need to actually compare the results across all the engines and in the end except in obvious cases, who is to say definitively what results are more relevant?
This much is safe to say though- while sometimes Cuil is reasonably accurate, there is definite room for improvement.
The founders of Cuil claim that rather than just primarilly looking at the number and quality of links to a webpage as Google’s technology does to determine ranking results, Cuil claims to understand more about the information on a page and the terms that people use to search.
Because of this fundamental shift in algorithm priorities, Cuil promises to re-emphasize the actual content of websites vs. in Googles case placing much too much emphasis on inbound, and especially non-reciprocal links to judge relevancy.
This is very ironic because it was precisely because of Search Engine Optimization professionals gaming the system by over-aggressive SEO (some would say) in the first place that led to Google adjusting thier algorithm to place more weight on inbound links to establish a so-called relevant site. In actuality however, while some SEO consultants definitely are mischievious and up to no good, the bulk of SEO companies are merely determining what best facilitates the search engine spiders ability to read and judge the content in a viable and effective way and then presenting it to them as such. That Google was incapable of weeding out over-optimized sites and not truly relevant sites is not necessarilly the fault of any SEO company, but rather falls on the shoulders of the “Big G” themselves for not having a more robust algorithm to detect this foul play, so it will be very interesting to see if Cuil’s algorithm over time can perform and retrieve the most relevant results based of the content on a given webpage rather than simply judging it to be the best largely because of the number of inbound links the site in question has.
Anyone that uses a search engine is painstakingly familiar with Wikipedia dominating the SERP’s with often times completely irrelevant pages merely because they have the most inbound links. After all, which is more preferrable as a result for your search engine query; a single page on a subject, or an entire website devoted to that same subject as the case may be? Sure, Wikipedia pages are sometimes very helpful and relevant but just as often the results return an outdated, or innacurate page, sometimes with information taken from the “authority site” that it actually outranking. Hardly the most relevant result many times, but due to the overwhelming number of inbound and unreciprocated links Wikipedia has (due to thier use of the no-follow attribute), Google judges it to be the best result based seemingly almost solely on that single factor.
Another key difference between Cuil and the other search engines is that Cuil does not log your I.P. information as does Google which is sure to be viewed as a positive by privacy activists.
As for cosmetic issues, one thing that seems to be universally condemned by just about everyone is Cuil’s odd choice of a 3 column design for it’s results as it makes it somewhat confusing as to which site actually ranks where although one does have the option to change to a 2 column results page which in this writers opinion is not much better.
Also, for a company that raised 33 Million in venture capital, one would think they could come up with a better domain name than “cuil” (pronounced as “cool”) which supposedly means “knowledge” in Gaelic.
So in summary, what is the final verdict on Cuil? Well honestly, after only 2 weeks it is still much too soon to tell. As mentioned earlier many of the results are irrelevant, with some being dead-on. Given the state of search today and with many people becoming more and more uncomfortable with the behemoth that is Google and competitors like Yahoo turning over thier users search data to the Chinese government, many people are hungry for a young upstart to challenge Goliath and the narrative of it being done by ex-Google employees defecting to start thier own operation has a nice ring to it, but only time will tell if they are up to the challenge.
Realistically though, Google is not going anywhere anytime soon if at all, however from what I have been able to see, Cuil’s results are already better than most “second tier” search engines (the ones below Google, Yahoo and MSN) so maybe given some time it’s not unreasonable to think Cuil can nibble away at some of Yahoo’s and MSN’s share of the pie… or they could simply agree to be bought out to avoid the competition. At this point it’s a crapshoot to know how things will shake out, but either way it will be interesting to watch.
Source: Professional Seo
suite, which also includes the Elastic Compute Cloud (EC2) for cloud processing power and Simple Storage Service (S3) for cloud storage. The Amazon triumvirate of SimpleDB, S3, and EC2 is meant solve the scalability needs of web developers with a utility-like model. Customers pay for just the storage, computations, and bandwidth they need, and none they don’t. While Google has yet to announce the pricing for BigTable, we presume it will share the same model as AWS.
