Tuesday, July 29, 2008

Challenging Google

I have special interest in web search, as I strongly believe that they have and will continue to have a major and increasing impact on our life and work. Google, of course, continues to lead and have the lion's share of the market. But like any industry, it would be healthy if strong competition emerges and contributes to innovation in search. This is not the case now, although you hear on an almost daily basis about a challenger coming into the picture and claiming to be doing what Google can't, and that it would become the new search standard. But that fizzles out fairly quickly and Google remains unshakable.

The last serious contender I remember was Jimmy Wales, the founder of Wikipedia, when declared that "the search is broken", and the he is about to fix it with his Wikia Search project. That was bout two years ago and we are yet to see anything serious coming out of that.

In the past few days, a new search engine named Cuil was launched, again to the tunes of grandiose claims that it is bigger, faster, and better than Google. The founding husband-and-wife team of former Google search architect Anna Patterson and Stanford professor Tom Costello, however, served better to show us how not to launch a search engine!

As it is, Cuil still have ways to go. I believe it was fairly self destructive to launch it the way they did. There should have been a soft launch, many months of low-profile alpha, a slow move to beta, and so forth, in a careful, gradual manner.

Challenging Google shouldn't be taken that lightly! Everybody knows that web search is a very serious undertaking, otherwise everybody would have done it. Serious in all aspects, including infrastructure scalability, robust ranking algorithms (no excuse that the server in charge of sports shut down!), and mostly results that make sense, plain and simple.

I am surprised that Cuil didn't approach it this way, as it sounds like a very common sense approach!

One example of a basic expectation not fulfilled is when I searched for my company, iHorizons. For years, when I search Google for iHorizons, not only it places our website at the first place in its results, it also suggests to auto-complete the name correctly while I type it. And the rest of the results are mostly meaningful and diversified. Cuil, on the other hand, didn't have our website in the first few pages of their results, and the rest of entries found are not so useful; mostly news for a small number of sources.

So we still have ways to go to have a serious competitor to Google. And again, if Google remains the absolutely dominant search portal the way it is today, I expect innovation to slow down, at a time when we really need a lot of advancement to make search easier and more useful.

Having said that, it is important to remember that Google is not the dominant search portal in all countries. Far from it. Google takes a back seat to local search engines in some of the world's most-wired countries, like Korea, where Google has a mere 1.7 percent share of Korean Web searches, while the local search engine Naver handles more than 77 percent of Korean search. People there 'Naver' information rather than 'Googling' it. Same situation in China, where Baidu has the top spot, owning 64% of the market, compared with Google's 26%. In Russia people "Yandex" for information. I will elaborate on this in a future blog, with particular focus on Arabic search, and why we don't have a regional engine that has the lead in Arabic search. Arabic search is very broken, but who is going to fix it?