VentureBeat, a site that covers venture funding is covering a company called Powerset that is supposedly getting its hands on a new technology that will let it understand contextual meaning in search phrases, something Google has been trying to do for years. A crazy claim is being made that this technology will allow the company to surpass Google someday. Are they nuts? Have they actually sat down and done the numbers?
Even if the technology isn’t PR bullshit, it’s still 10 years too late. How can a yet un-invented technology be too late you ask?
Because searching has become standardized as a keyword game. When the web first started out, people typed in stuff like “where is the airport?” But nowadays, people will type in “airport location”, which works great. People no longer type full sentences into queries, and that’s a good thing. The web is going onto mobile devices at an increasing rate. This means people are even less likely than ever to want to type out an entire sentence or phrase. It’s all about keywords now.
This company and its venture capitalists misunderstands the race. To fight Google, you can’t just have the best results by a small margin; your results have to smash Google’s results into tiny little pieces. And to do that, you first need to index at least 10 billion websites just to catch up to Google, an effort that takes years. To catch Google, you have to also build a sustainable business around supporting the hundreds of thousands of servers required to crawl and store the entire web. You have to convince people that the extra effort of typing more words is worth it even if your index crawls less pages, which increases the likelihood they won’t find what they want.
Had this technology come out in the days of Altavista, this company might be rolling in the green. But to complicate things further, in the era of Google, the race is no longer just about search. Google provides email, chat, calendar, and word processor services. Google has groups, video, and most importantly the largest and best funded online ad and publisher network on the planet.
The biggest reason why this technology is late is because of the time it will take to overcome Google. Even if a company was truly better than Google, it would take it years just to catch up. Microsoft has thrown billions at the search problem and they are actually losing share to Google and Yahoo. Even though nearly every computer in the world have MSN as their default home page, people actually type in “Google.com” and search from there. Google is that entrenched. It’s even in the dictionary now!
And even with the vastly superior search, the new competitor would need to fight Microsoft and Yahoo, both which are companies that own the top two websites in the world. The fact that these two companies have immensely popular portals is one of the big reasons they can still compete with Google.
I’m not saying fighting Google is impossible. But it’s improbable, especially if you are a no-name company. Talking about a yet to be officially announced technology as if it is guaranteeing this unknown company a top spot is just a grab at publicity. Any realistic analysis would show that unseating Google would take billions of dollars in marketing, hundreds of millions of dollars in infrastructure, and a general assumption that Google, Yahoo, or Microsoft will stand still.
I’d love to see them try. But even if my friend told me, “Michi, I try this cool search engine that understands what you say,” I still wouldn’t switch. In a nut shell, I won’t switch search engines for that 0.1% time where I can’t find what I need. Neither would most people.