Meanwhile, BackRub-turned-Google was growing to the point where it was difficult to run using Stanford’s facilities. It was becoming less a research project than an Internet start-up run from a private university. Page and Brin’s reluctance to write a paper about their work had become notorious in the department. “People were saying, ‘Why is this so secret? This is an academic project, we should be able to know how it worked,’” says Terry Winograd.
Page, it seemed, had a conflict about information. On one hand, he subscribed heartily to the hacker philosophy of shared knowledge. That was part of what his project was all about: making human knowledge accessible, making the world a better place. But he also had a strong sense of protecting his hard-won proprietary information. He remembered Nikola Tesla, who had died in poverty even as his inventions enriched others. Later, there would be speculation whether Page, a private person to begin with, had pulled back a little more after his father’s death in June 1996. Scott Hassan recalls that the team conveyed its condolences to Page that month, but Hassan didn’t speak much about the loss with Page. “Mostly we talked about technical stuff,” he would recall. Mike Moritz, one of the venture capitalists who would fund Google, later surmised that “a large part” of Page’s later wariness could be associated with that loss. “He felt that the world was pulled out from underneath him,” Moritz said. “It makes it hard to trust anything again.”
But it wasn’t just the secrecy that stalled Brin and Page. Writing a paper wasn’t as interesting to them as building something. “Inherently, Larry and Sergey aren’t paper-oriented—they’re product-oriented,” says Winograd. “If they have another ten minutes, they want to make something better. They don’t want to take ten minutes to tell you something they did.” But finally Winograd convinced them to explain PageRank in a public forum. They presented a paper called “The Anatomy of a Large-Scale Hypertextual Web Search Engine” at a conference in Australia in May 1998.
Arthur Clarke once remarked that the best technology was indistinguishable from magic. The geeks of Silicon Valley, assuming he was talking about them, have never forgotten that and have invoked the quote in countless press releases about their creations. But Google search really did feel like magic. At Stanford, Larry’s and Sergey’s professors and friends were using the search engine to answer questions and telling their friends about it. Google was handling as many as 10,000 queries a day. At times it was consuming half of Stanford’s Internet capacity. Its appetite for equipment and bandwidth was voracious. “We just begged and borrowed,” says Page. “There were tons of computers around, and we managed to get some.” Page’s dorm room was essentially Google’s operations center, with a motley assortment of computers from various manufacturers stuffed into a homemade version of a server rack—a storage cabinet made of Legos. Larry and Sergey would hang around the loading dock to see who on campus was getting computers—companies like Intel and Sun gave lots of free machines to Stanford to curry favor with employees of the future—and then the pair would ask the recipients if they could share some of the bounty.
That still wasn’t enough. To store the millions of pages they had crawled, the pair had to buy their own high-capacity disk drives. Page, who had a talent for squeezing the most out of a buck, found a place that sold refurbished disks at prices so low—a tenth of the original cost—that something was clearly wrong with them. “I did the research and figured out that they were okay as long as you replaced the [disk] operating system,” he says. “We got 120 drives, about nine gigs each. So it was about a terabyte of space.” It was an approach that Google would later adopt in building infrastructure at low cost.
Larry and Sergey would be sitting by the monitor, watching the queries—at peak times, there would be a new one every second—and it would be clear that they’d need even more equipment. What next? they’d ask themselves. Maybe this is real.
Stanford wasn’t kicking them out—the complications of running the nascent Google were outweighed by pride that something interesting was brewing in the department. “It wasn’t like our lights were dimming when they would run the crawler,” says Garcia-Molina, who was still hoping that Larry and Sergey would develop their work academically. “I think it would have made a great thesis,” he says. “I think their families were behind them to get PhDs, too. But doing a company became too much of an attraction.”
There was no alternative; no one would pay enough for Google. And the happy visitors they were attracting gave them confidence that their efforts could make a difference. After years of dreaming how his ideas could change the world, Larry Page realized that he’d done something that might do just that. “If the company failed, too bad,” says Page. “We were really going to be able to do something that mattered.”
They went back to Dave Cheriton, who encouraged them to just get going. “Money shouldn’t be a problem,” he said. Cheriton suggested that they meet with Andy Bechtolsheim. Brin dashed off an email to Bechtolsheim that evening around midnight and got an immediate reply asking if the two students could show up at eight the next morning at Cheriton’s house, which was on the route Bechtolsheim used to go to work each day. At that ungodly hour Page and Brin demoed their search engine for Bechtolsheim on Cheriton’s porch, which had an ethernet connection. Bechtolsheim, impressed but eager to get to the office, cut the meeting short by offering to write the duo a $100,000 check.
“We don’t have a bank account yet,” said Brin.
“Deposit it when you get one,” said Bechtolsheim, who raced off in his Porsche. With as little fanfare as if he were grabbing a latte on the way to work, he had just invested in an enterprise that would change the way the world accessed information. Brin and Page celebrated with a Burger King breakfast. The check remained in Page’s dorm room for a month.
Soon afterward, Bechtolsheim was joined by other angel investors, including Dave Cheriton. One was a Silicon Valley entrepreneur named Ram Shriram, whose own company had recently been purchased by Amazon.com. Shriram had met Brin and Page in February 1998; although he had been skeptical about a business model for search engines, he was so impressed with Google that he had been advising them. After the Bechtolsheim meeting, Shriram invited them to his house to meet his boss Jeff Bezos, who was enthralled with their passion and “healthy stubbornness,” as they explained why they would never put display ads on their home page. Bezos joined Bechtolsheim, Cheriton, and Shriram as investors, making for a total of a million dollars of angel money.
On September 4, 1998, Page and Brin filed for incorporation and finally moved off campus. Sergey’s girlfriend at the time was friendly with a manager at Intel named Susan Wojcicki, who had just purchased a house on Santa Margarita Street in Menlo Park with her husband for $615,000. To help meet the mortgage, the couple charged Google $1,700 a month to rent the garage and several rooms in the house. At that point they’d taken on their first employee, fellow Stanford student Craig Silverstein. He’d originally connected with them by offering to show them a way to compress all the crawled links so they could be stored in memory and run faster. (“It was basically to get my foot in the door,” he says.) They also hired an office manager. But almost as if they were still hedging on their PhDs, they maintained a presence at Stanford that fall, coteaching a course, CS 349, “Data Mining, Search, and the World Wide Web,” which met twice a week that semester. Brin and Page announced it as a “project class” in which the students would work with the repository of 25 million web pages that they had captured as part of what was now a private company. They even had a research assistant. The first assigned reading was their own paper, but later in the semester a class was devoted to a comparison of PageRank and Kleinberg’s work.
In December, after the final projects were due, Page emailed the students a party invitation that also marked a milestone: “The Stanford Research Project is now Google.com: The Next Generation Internet Search Company.”
“Dress is Tiki Lounge wear,” the invitation read, “and bring something f
or the hot tub.”
2
“We want Google to be as smart as you.”
Larry Page did not want to be Tesla’d. Google had quickly become a darling of everyone who used it to search the net. But at first so had AltaVista, and that search engine had failed to improve. How was Google, led by two talented but inexperienced youngsters, going to tackle the devilishly difficult problems of improving its service?
“If we aren’t a lot better next year, we will already be forgotten,” Page said to one of the first reporters visiting the company.
The web was growing like digital kudzu. People were coming to Google in droves. Google’s plan was to get even more traffic. “When we started the company, we had two computers,” says Craig Silverstein. “One was the web service, and one was doing everything else—the page rank, the searches. And there was a giant chain of disks that went off the back of the computer that stored twenty-five million web pages. Obviously that was not going to scale very well.” Getting more computers was no problem. Google needed brainpower, especially since Brin and Page had reached the limits of what they could do in writing the software that would enable the search engine to grow and improve. “Coding is not where their interests are,” says Silverstein.
The founders also knew that Google had to be a lot smarter to keep satisfying users—and to fulfill the world-changing ambitions of its founders. “We don’t always produce what people want,” Page explained in Google’s early days. “It’s really difficult. To do that you have to be smart—you have to understand everything in the world. In computer science, we call that artificial intelligence.”
Brin chimed in. “We want Google to be as smart as you—you should be getting an answer the minute you think of it.”
“The ultimate search engine,” said Page. “We’re a long way from that.”
Page and Brin both held a core belief that the success of their company would hinge on having world-class engineers and scientists committed to their ambitious vision. Page believed that technology companies can thrive only by “an understanding of engineering at the highest level.” Somehow Page and Brin had to identify such a group and impress them enough to have them sign on to a small start-up. Oh, and they had a policy that limited the field: no creeps. They were already thinking of the culture of their company and making sure that their hires would show traits of hard-core wizardry, user focus, and starry-eyed idealism.
“We just hired people like us,” says Page.
Some of Google’s early hires were simply brainy recent grads, people like Marissa Mayer, a hard-driving math whiz and ballet dancer in her high school in Wausau, Wisconsin, who had become an artificial intelligence star at Stanford. (During her interview with Silverstein, she was asked for three things Google could do better; ten years later, she was still kicking herself that she listed only two.) But Page and Brin also went after people with résumés more often seen in the recruitment offices of Microsoft Research or Carnegie Mellon’s CS department. One of their first coups was a professor at the University of California at Santa Barbara named Urs Hölzle. He’d played with the earlier crop of search engines such as AltaVista and Inktomi and concluded that, as a computer scientist familiar with Boolean syntax and other techniques, he could use those techniques to find what he wanted on the Internet. But he assumed that search would never be something his mother would use. Google instantly changed his mind about that: you just typed in what you wanted, and, bang, the first thing was right. Mom would like that! “They definitely seemed to know what they were doing,” he says of Larry and Sergey.
More important to him, when he visited the new company in early 1999, he understood that though he had no background in information retrieval, the problems Brin and Page were working on had a lot in common with his own work in big computer systems. This little search engine was butting up against issues in performance and scalability that only huge projects had previously grappled with. That was Google’s secret weapon to lure world-class computer scientists: in a world where corporate research labs were shutting down, this small start-up offered an opportunity to break ground in computer science.
Hölzle, still wary, accepted the offer but kept his position at UCSB by taking a yearlong leave. He would never return. In April he arrived at Google with Yoshka, a big floppy Leonberger dog, in tow, and dived right in to help shore up Google’s overwhelmed infrastructure. (By then Google had moved from Wojcicki’s Menlo Park house to a second-floor office over a bicycle shop in downtown Palo Alto.) Though Google had a hundred computers at that point—it was buying them as quickly as it could—it could not handle the load of queries. Hundreds of thousands of queries a day were coming in.
The average search at that time, Hölzle recalls, took three and a half seconds. Considering that speed was one of the core values of Page and Brin—it was like motherhood, and scale was apple pie—this was a source of distress for the founders. “Basically during the middle of the day we were maxed out,” says Hölzle. “Nothing was happening for some users, because it would just never get a page basically back. It was all about scalability, performance improvements.” Part of the problem was that Page and Brin had written the system in what Hölzle calls “university code,” a nice way of saying amateurish. “The web server couldn’t handle more than ten requests or so a second because it was written in Python, which is a great idea for a research system, but it’s not a high-performance solution,” he says. He immediately set about rewriting the code.
Hölzle was joined by other computer scientists who were more daring in taking the leap to permanent Google employment. This included a minimigration of engineers from DEC’s research division. Established legend in Silicon Valley cited Xerox’s Palo Alto Research Center (PARC) as the canonical lab brimming with breakthrough innovation that had been misunderstood, buried, or otherwise fumbled by the clueless parent company. (Its inventions included the modern computer interface with windows and file folders.) But when it came to missed opportunities, PARC had nothing on DEC’s Western Research Laboratory, which was handed over to Compaq when that personal computer company bought Digital Equipment Corporation in 1998. (In 2002, Hewlett-Packard would acquire Compaq.) In 1998, two years before Apple even began work on the iPod, DEC engineers were developing a digital music player that could store a whole music collection and fit in your pocket. In addition, DEC had some of the founding fathers of the Internet, as well as scientists writing pioneering papers on network theory. But DEC never used its engineers’ ideas to help AltaVista become Google. (“From the moment I left DEC, I never used AltaVista,” says Louis Monier, who split in 1998. “It was just pathetic. It was completely obvious that Google was better.”) So it was little wonder that some of them went to Google. “The number [of former DEC scientists at Google] is really kind of staggering,” says Bill Weihl, a DEC refugee who came to the company in 2004.
One of the DEC engineers had already independently discovered the power of web links in search. Jeffrey Dean suspected that it would be helpful to web users if a software program could point them to pages that were related to the ones they liked. In his vision, you would be reading an article in The New York Times and his program would pop up, asking if you’d like to see ten other interesting pages related to the one you were reading.
Dean had never been much interested in information retrieval. Now that he suspected a revolution was afoot, he was. But his attempts to join up with the AltaVista crew ended ignominiously. “The AltaVista team had grown really fast,” he says, “and hired a bunch of people who I think were not as technically good as they could have been.” In other words—get me away from here. In February 1999, Dean bailed from DEC to join a start-up called mySimon.
Within a few months, though, he was bored. Then he heard that Urs Hölzle, whom he’d known through his grad school adviser, had joined up with the guys who did PageRank. “I figured Google would be better because I knew more of the people there, and they seemed like they were more technically savvy,??
? he says. He was so excited about working there that even though his official starting date wasn’t until August 1999, in July he began coming to Google after his workday at mySimon ended.
Dean’s hiring got the attention of another DEC researcher, Krishna Bharat. He had also been thinking of ways to get web search results from links. Bharat was working on something called the Hilltop algorithm, which algorithmically identified “expert sites” and used those to point to the most relevant results. It was something like Jon Kleinberg’s hub approach, but instead of using AltaVista as a prewash to get top search results and then figure out who the expert sites were, Bharat went straight to a representation of the web—links and some bits from the pages—stored in computer memory. Bharat’s algorithms would roam around the “neighborhood of the query” to find the key sites.