A search engine which has been making waves in the blogosphere for the last day or two is Cuil (pronounced “cool”). You’ve probably used it twice already, and its random search results (in some cases, the worst ever served by a search engine) do less to serve and more to amaze. I made up my mind this morning to not write about Cuil since I think it had been written about a little too much, but now it hardly seems like it matters.
For you slowpokes, the back story of Cuil is that it was started by a couple ex-Googlers who played a crucial part in what Google search is today, they raised $30 million for the startup and brought on board more Googlers, who have then built a search engine with an index of 125 billion pages (supposedly more than Google’s) and close to 1% relevancy, though no one is judging since it is, literally, their first day on the job.
I think what’s more interesting than the search engine itself, which is like any other “kill Google” new-search-engine-on-the-block, is the question of whether we need another Google, and if it is theoritically possible for someone to create a search engine which has results 25% – 50% better than Google’s and get 1 billion people to switch to it.
Robert Scoble doesn’t think so. He argues, “I still doubt many people would switch. Why? Distribution. Huh? Well, my Firefox browser has Google built into it. Most people have no idea how to switch it. Most people, on our tests, really don’t understand much of anything except that that little box probably now goes to Google. The Google. It’s so pervasive of an expectation at this point that many people type URLs into that box. Or, type the word “Yahoo” into that box so they can get to their email and other Yahoo services. Is Cuil going to be able to get into this game? No way, no how.”
I think Scoble is right on the dot, as close-minded as it sounds. Cuil’s problem isn’t relevancy, even given that it has cracked it. Cuil’s problem is distribution. Mass, billion user-level distribution. We like to think that the market place will at all costs reward anything which is better than something else, but I simply don’t think this will be the case should a real Google-killer arrive (and Cuil isn’t it). Instead, here is how I think it will more likely happen:
- Google killer arrives.
- Early adopters try it out, like it, have luck with it. Love it.
- Google killer gets bought by Microsoft and is never to be heard of again, found later as a part of Live.
And that’s the sad and depressing fact of being a 21st century Google-killer. See, when Google took over AltaVista, the situation was very different. Unlike Google, AltaVista sucked. They had cracked probably a 25% relevancy rate, which is close to being the only market option at the time. Google did 50%, eventually people noticed it, and AltaVista is not to be heard of today.
Which brings me to Cuil (or any Google killer’s) problem: Google today is probably about 90% relevant. This means it satisfies probably 90% of people’s searches on the Internet, that is of course, given the appropriate and relevant pages exist. This means, though you may have better search results than Google, Google is going to work 90% of the time regardless of its second-grade search results, and as a result, 90% of the people on this planet have no reason to care about you.
So let’s say a search engine cracks 100%. The result: as long as Google’s relevancy doesn’t degrade over time, or they are overwhelmed somehow by spam overnight, Google ain’t moving. 70%+ search dominance, we’re still alive! This is because of one simple reason: distribution. Google has manage to crack distribution, branding, presence, while being “just good enough,” and this is all the world asks from a search engine — no 100% relevancy, nothing else.
But that is not to say that all hope is lost for aspiring Google killers. I think if you can crack, firstly, 90%+ relevancy, and secondly, a user interface which is as close to what Google’s own was at its time, which says a lot seeing that it’s remained the same for 10 years, and if you can resist the temptation of a $100 million – $400 million buyout from Steve Ballmer, you may have something in your hands. And then, it would be a long wait — maybe years — till you can get even a fraction of the distribution and scaling that Google has, but after that, you will have a Google killer in your hands.
The chances of that happening of course is very unlikely. What is more likely is a growing pool of aspiring Google killers in the market, and, as a result, a growing pool of possible acquisitions for Microsoft to choose from. My prediction: Microsoft gobbles gobbles gobbles through this growing pool, and eventually Live Search is as good as Google’s, but distribution is yet to be cracked. In other words: Google remains king.