3by9

Nobody Will Topple Google Search

In last week’s podcast we talked a little about company lifecycles and how a company’s dominance will ebb and flow, with competitors sprouting up taking advantage of the veteran company’s weaknesses. I made the bold statement that Google’s dominance in search can never be beaten, so here are my expanded thoughts on that.

Scale

A few years ago I read a technical paper written by a Google engineer, and in that paper it was revealed that at any given time, Google has a full cache of the entire Internet in RAM of its server array. That’s how it calculates Pagerank so quickly, giving results for any query on the planet in an instant. The Pagerank algorithm is based on many factors, one of which is the recursive analysis of links pointing into a particular site, all of the links linking into that linking site, and then on down the line. The Google platform is made up of over a half million of cheap PCs, where their power lies not in their processor but in the customized operating systems and distributed software that they all run. No one has a larger cache of Internet pages than Google, and no one offers a search mechanism that’s faster than Google. If a company started fresh and dedicated all their time to producing a more flexible and efficient network of computers, it could take a few years (and perhaps hundreds of millions of dollars) to do it, and by then Google would be a few years ahead of them since they’re growing their server network and improving their platform server every day. Since we’re talking about search here, and “better search” means 1) more searchable pages, and 2) a better/faster algorithm to find data, the sheer size of Google’s computing facilities is daunting and insurmountable.

Quality

Since no search engine can compete with Google on quantity of searchable pages, they resort to competing on search quality. Ask’s new advertising campaign focuses on their new and improved algorithm that purportedly offers “better” results than Google. The problem is that this idea of “better” cannot really be defined since people are generally happy with Google’s search results (if they weren’t, then they’d go somewhere else, and few people do) so it’s difficult to put a finger on what would make them better, if anything. Companies have experimented with adding images and videos to search results, but the trouble is that if anyone comes out with a good idea, Google can just add the same feature and there goes the competitive advantage. When someone is searching for something on the Web, they generally don’t read past the first few results, and the problem for Google competitors is that Google’s first page of results is very good.

Mahalo is a new company that is attacking search from a different perspective, and they have their own take on what “quality” results are. They have thousands of human-edited “results pages” for users’ most-searched items, and have a lot of good meta information on these pages so they’re more than just a list of links. Jason Calacanis is going after Google by trying to boost the quality of search results, but the problem is that no one will find these gems of pages if Google doesn’t rank them high enough on their own search engine.

Resources

To compete with Google you need billions of dollars and the smartest engineers on the planet. Only a few companies have this (Microsoft, Yahoo!, etc.) and they haven’t taken down the giant yet. Microsoft isn’t focusing much on search these days, but Yahoo! still is, ever since Google overtook them in the search world. Yahoo! is splintered at its core, with so many different services and content sites that it seems they’re not able to pull all their resources into one specific thing in order to excel. That’s the trouble with competing with Google, search is their core service that powers all their revenue (search for a term, get AdWords; click on a page, see AdSense) so they will never be short on resources devoted to search. Their search algorithm and the architecture on which it runs is their bread and butter, and everyone knows it. Microsoft is a software company, Yahoo! is a content portal (or something, what is Yahoo… really?), and Google is search. A company would have to put the full weight of its resources into developing search for them to be competitive, and even then it should be known that Google has over a decade of search technology under its belt and has a monster lead in the industry, and they’re still growing.

The only reason why Google search’s dominance could wane is if they self-destruct internally and cause it to happen.