Fighting to stay No. 1

Published: Sunday, October 1, 2006 at 6:01 a.m.
Last Modified: Sunday, October 1, 2006 at 12:00 a.m.
With its heft, power and sizzle among everyone from term-paper researchers to Wall Street financiers, Google has made an astonishing ascent from little-known start-up to Internet icon.
But even with its stratospheric market value of $123 billion, Google finds itself in an odd position these days: always looking over its shoulder for the next threatening rival.
A rare extended look inside the company's Googleplex headquarters strips away the recent mythologizing of Google to reveal a different side: the manic scramble to invent breakthrough technologies, or risk becoming the latest dot-com also-ran.
In the conference room where Google executive Marissa Mayer holds court, time is precious as she considers ideas to refashion the famous, minimalist look of Google's search pages. Indeed, a large digital clock on the wall ruthlessly counts down the seconds.
"You should have a giant 'Click Me' button. You need something that screams 'Please, Click Me,' " she says at one point. To another proposal: "We have one shot." Then, "Next."
Fast action like that has helped Google stay atop the frenzied world of the Internet while feeding an appetite for expansion nearly as vast as the Web itself. Just eight years after two Stanford University graduate students created the strikingly fast and comprehensive search engine, Google wants every move on the Internet to go through its disarmingly simple home page, period.
But Google is finding it impossible to stay on top of everything, everywhere. In fact, it isn't the leader in the latest rush for Internet gold: video over the Internet. In that lucrative domain, a tiny company called YouTube rules.
Google's poor showing in video highlights an often-ignored fact: While the company started with a run of blockbuster hits - search, maps and Gmail - it hasn't had one since.
"The sense that Google can do everything well has been tarnished," said Joe Kraus, who co-founded and then sold, which pioneered computer searches in the early 1990s. "They've made people say, 'Wait a second, not everything coming out of Google is amazing."'
The trouble is, Google's phenomenal stock price, topping $400 a share, is based on an assumption that the company will continue to amaze.
Google's struggle to catch up was center stage this summer when a design team crowded together in a Googleplex conference room to upgrade the firm's video service. The designers tried to keep their eyes focused on Google, but every decision inevitably turned to YouTube.
One engineer wanted a new "Play" page to pop up when users return to Google Video. "That was one thing YouTube didn't do," he said brightly.
But they couldn't escape a nagging concern. "We look just like YouTube," said a leader of the team, Peter Chane, group product manager of Google Video. "We're a cleaner, more thoughtful YouTube, (only) with less features."
Internet video started as the outpouring of made-from-scratch creativity by citizens with video cameras. Now, though, the race between YouTube and Google is getting increasingly commercial.
YouTube in late June inked a pact with NBC to feature previews of the network's programs on YouTube, while NBC promised to promote YouTube on popular NBC shows such as "The Office." And it signed a deal this month to show videos from Warner Music and share in the advertising revenues the videos generate.
Google recently has cut deals with some of the hippest names in media, including MySpace and MTV. In the MySpace agreement, Google is paying the social-networking site up to $900 million to place its search page on MySpace, which many young music lovers use to view videos of their favorite garage bands.
YouTube co-founder Steve Chen, 27, a University of Illinois graduate, relishes watching Google try to catch up. He thinks YouTube has one key advantage.
"Every person that we're hiring, every person we're bringing on the team, is focused on video," said Chen.
When fellow grad students Larry Page and Sergey Brin founded Google, they vowed to let data drive every decision. One key edict requires that, across the enterprise, people spend 70 percent of their time on Google's core search and advertising businesses, 20 percent on other existing products and just 10 percent on new initiatives.
The 70-20-10 rule reflects the Google founders' understanding that search is their key franchise. But in the course of Google's meteoric growth, from its start as an unproven search engine in 1998 to last year's $6.1 billion in revenues, the company's idiosyncratic orthodoxies are proving hard to enforce.
Eric Schmidt, Google's chief executive, requested an audit of employees' actual work habits recently, and found they were spending only 60 percent of their time on Google's core business. Even more worrisome to Schmidt: Google had no idea it had veered so far off course.
"It was quite alarming to find that we were below the 70 percent without knowing it," said Schmidt, a Sun Microsystems veteran recruited by Page and Brin in 2001. "It's a real crisis, in the sense that we are not fully deploying people on the right things."
Schmidt said he ordered up new internal audits to monitor how employees spend their work time, and the company has since corrected that flaw.
At the heart of Google's business is something most people learned, and quickly forgot, in high school: a mathematical algorithm. Google's algorithm is the series of instructions that provides the electronic flowchart needed to tell the synapses in Google's brain how to work.
Central to the algorithm is PageRank, a patented formula that helps order results based largely on popularity. If many sites with high PageRank link to a Web page, then it, too, receives a high PageRank.
Google co-founder Page put it this way: "You're important if other people think you're important."
The system also examines about 200 other signals, ranging from how often keywords appear on a page to their proximity to each other. "John F. Kennedy" appearing next to "biography" has a different value than a document where those two phrases are paragraphs apart.
Google deploys automated programs called Googlebots to scan the Web for everything it can find and compile the results into an index. Google claims its index contains three times as many documents stored as any competitors, although there is no way to independently verify that.
The size of Google's index makes it particularly good at delivering results for obscure queries. But competitors say that's not what really matters.
"If you have a rare query, go to Google," said Steve Berkowitz, who left a job as CEO of the search engine to become Microsoft's senior vice president of online services. "They can have that 3 percent, and I'll take the other 97 percent."
At Yahoo, meanwhile, software engineers regularly test sets of tens of thousands of randomly selected query terms. After those queries are run, editors mark up what results are useful or not. Then Yahoo's programmers write commands that help its search engine reflect the judgment of the editors.
Yahoo believes its approach delivers better "perceived comprehensiveness," even though its index is smaller than Google's.
When Hurricane Katrina was driving toward the Gulf Coast last September, a "Katrina" query on Google turned up a Web designer's site,, that had nothing to do with the storm. Yahoo programmers already had manually inserted a fix that drove users to information about the impending disaster.
"We think that (the human touch) definitely adds to the product, as opposed to the kind of a purist view that you have to have an algorithm that necessarily does these things," Pedersen said.
Still, success doesn't last forever in Silicon Valley. The Googleplex itself is testament to that.
The glimmering glass-and-steel complex wasn't built by Google. A company called Silicon Graphics built it - then went bust.

Reader comments posted to this article may be published in our print edition. All rights reserved. This copyrighted material may not be re-published without permission. Links are encouraged.

Comments are currently unavailable on this article

▲ Return to Top