Google didn’t take pages from the established Silicon Valley playbook because, in a way, they had never bought into it. They didn’t try to Get Big Fast. Instead, Page and Brin were almost manically focused on endlessly iterating and improving upon their Big Idea, making sure it was the most comprehensive, reliable and—most important—speedy search engine in the world. Nothing Google did in its first years distracted the company from improving on its core product. This confidence that they could do everything better proved, in the coming years, to be something of Google’s secret sauce. Not only did Google’s search engine continue to be superior to any rival in existence, it slowly but surely widened the gap between its version of search and the competition. And their frugality paid off in efficiency. Some observers estimated that “for every dollar spent, Google had three times more computing power than its competitors.”14
Frugality and efficiency were not just virtues, they were also philosophical and aesthetic differentiators. Google’s home page was simply the Google logo, a text field to enter a search query, a search button to execute that query and a button that said I’M FEELING LUCKY, which automatically took you to the first result returned. If you went to the search results page, you only got a list of links. And that was it. No ads, no banners, no weather, no stock quotes, no horoscopes. All the rest of the page was just copious white space. In an age of portals where every other search site was a sea of distractions meant to keep you from, you know, getting to the page you were looking for, Google stood out from the crowd with its single-minded purpose and simplicity. By keeping the pages to almost exclusively text, Page and Brin could ensure they loaded quicker than the search pages of their competitors, and expensive processing power wasn’t wasted loading graphics.
This all paid dividends many times over in Google’s steady growth. By 1999, usage of the search engine was increasing by as much as 50% a month.15 From 100,000 searches a day at the beginning of that year, Google searches grew to an average of 7 million per day by the end of it.16 Overall traffic to the Google homepage was peanuts compared to the numbers a site like Yahoo was pulling down, but in the case of Google, its users came via word of mouth alone. Not a dime was spent on marketing or promotion. Rave reviews from the media continued to turn people on to the service. The New Yorker said Google was “the default search engine of the digital in-crowd.”17 Time Digital said: “Google is to its competitors as a laser is to a blunt stick.”18 Ordinary users simply told one another about how great and useful Google was. More often than not, users would become Google converts for life.
An early article on Google in Fortune from November 1999 summed up a new user’s experience. Describing the site as “inscrutable magic,” journalist David Kirkpatrick offered this anecdote. On the occasion of the 1999 American League playoffs, Kirkpatrick typed “New York Yankees 1999 playoffs” into both Google and Alta Vista. “The first listing at Google took me directly to data about that night’s game,” Kirpatrick wrote. “The first two at Alta Vista linked to info about the 1998 World Series.” Only by clicking the third Alta Vista link, and then visiting an additional link, did he find the information he was originally searching for. Kirkpatrick’s conclusion: “Google really works.”19
In that same article, Sergey Brin was quoted as boasting, “We’re building a way to search human knowledge.” If Google was meant to organize all the information in the world, it would need resources on an industrial scale. That same brashness continued to manifest itself when Google needed to raise yet more money.
Despite the glut of search companies already on the market, Google had gotten the attention of venture capitalists, and they were ready to invest in these refugees from academia. But, confident as ever, Page and Brin gave off the impression that they didn’t need anyone’s help or money. In meetings with potential backers, the pair refused to divulge even basic details about how their service was operating. Their stonewalling even led one prominent VC to storm out of their office in anger. “Larry and Sergey didn’t have the language to say things nicely,” recalled Salar Kamangar, an early employee who bore witness to Google’s general evasiveness during the fundraising process. “They’d be kind of blunt and say, ‘We can’t tell you.’ And the VCs would get very frustrated.”20 The truth was, Page and Brin did not want to take money from just any old VC. They only wanted the best: Kleiner Perkins and Sequoia Capital. The pair proposed that each firm, the blue chips of Silicon Valley venture, take a coequal stake in Google. There was usually one “lead” investor in a round of startup financing, and both KP and Sequoia had enough clout on their own that they had never before deigned to share the spotlight with another firm.
Page and Brin wanted the firms to split the round because that would allow them, as the founders, to maintain a majority share in the company, and thereby retain control of their own destiny. They even had the temerity to issue an ultimatum: each firm would invest $12.5 million in Google, for a total of $25 million, take it or leave it. On June 7, 1999, the VCs took the deal, and Kleiner’s John Doerr and Sequoia’s Mike Moritz joined Google’s board of directors. The only concession the money men had been able to wring out of Page and Brin was a promise to hire someone experienced to take over as CEO of the company at some point in the near future.
This huge round of financing not only put Google firmly on the technology world’s map, it went a long way toward ensuring the company’s long-term survival. This war chest of money, coming just before the dot-com bubble burst, combined with Larry and Sergey’s frugal ways, meant that Google would survive the coming nuclear winter. Had Google waited a further year to raise money, it might not have been able to. And by virtue of being flush with cash when the rest of Silicon Valley was seemingly going belly-up, Google was able to have its pick of talent when the dot-com layoffs began.
Just as it had been frugal when others were profligate, Google also bucked prevailing dot-com habits when it came to hiring. The company put off drafting an army of sales and marketing people until much later. Instead, in 1999 and 2000, Google staffed up with—what else?—brainiacs. Larry and Sergey hired software engineers, hardware engineers, network engineers, mathematicians, even neurosurgeons. Just as with every other facet of their company, Page and Brin wanted only the very best. They wanted Ph.D.’s and scientists. Google would become notorious for the rigorous way it interviewed and screened potential hires—and for its exacting selectiveness. For many years, every new employee was personally vetted by Brin and Page themselves, who expected candidates to measure up to their own intellectual standard. “We just hired people like us,” Page said.21
Google was able to attract talent because it was nothing short of beloved in Silicon Valley. Here was an Internet company that had solved a universally recognized problem through smart thinking alone. This created a reputational halo that was only enhanced by Larry and Sergey’s increasingly bold and public enunciation of Google’s mission, which was eventually formalized as an attempt “to organize the world’s information and make it universally accessible and useful.” While so many dot-com companies claimed to be changing the world by offering dog food online, here was a company that truly seemed revolutionary in the most expansive sense of that word. “Ultimately I view Google as a way to augment your brain with the knowledge of the world,” Sergey Brin said.22 It helped that Google positioned itself as the anti–dot-com startup. Glitz, hype and excess were out; frugality, hard work and earnestness were in. And when Google came up with its famous motto (Don’t Be Evil) everyone in technology read between the lines and believed that Google was staking a claim to be the anti-Microsoft.
Google did pick up a few habits from its dot-com brethren, but in typical Larry and Sergey fashion, it did so with a twist. By the time Google moved to its first truly professional digs—an office park in Mountain View that would be dubbed the “Googleplex”—a system of perks for Google’s workers were put in place, but they were instituted with an eye toward productivity. The food in the cafeteria was al
ways free, with an in-house gourmet chef; private bus lines picked up workers from around the Valley to shuttle them to work; masseuses roamed the hallways; there were free fitness classes and gyms; and on and on. But every one of these perks was self-consciously provided as a way to keep workers motivated and productive. The free cafeteria meant that Google employees didn’t have to leave the office in the middle of the day and could get back to work with ease. In the bathroom stalls were quizzes and coding tips to help people stay sharp. The shuttle buses had WiFi on them, so employees could be productive on the way to and from the Googleplex. Healthy, clear-headed workers could do better coding, or so the thinking went.
All of this combined to make Google the technology company to join right as the dot-com bubble burst. If you got hired at Google, it elicited envy from your peers not only because they felt you were doing the most interesting work in technology, but because it meant you were among the best and the brightest. Anyone could get hired at a dot-com toward the end of the decade. But not everyone—even the smartest of the smart—could make the cut at Google. And when the bubble burst and it was seemingly the only company still hiring, the dream of the nineties was alive in the Googleplex.
■
GOOGLE HAD ALWAYS BEEN OBSESSED with its logs, the reams of data its users provided by their billions of searches. Google’s engineers used this data to improve the algorithms, but as the company was committed to “organizing the world’s information,” it also had a fascination with how search behavior revealed the world’s obsessions in real time. Eventually, products like Google Trends and Google Zeitgeist would allow us all to peek inside the planet’s collective unconscious, surfacing perennial obsessions like “sex” or “porn” but also faddish searches like “Paris Hilton” or “Justin Bieber.” In the year 2000, the hot search term was “MP3.” This was because, across the country, a teenager just barely into his first year of college had dreamed up a program that would break the Internet wide open just as definitively as Google’s algorithms were doing.
Shawn Fanning was a member of the first true web generation, born November 22, 1980, in the working-class Boston suburb of Brockton, Massachusetts. Earlier than most people his age, Shawn became a heavy user of online chat, especially Internet Relay Chat. It was on IRC that Shawn Fanning fell deeply in with the teenage hacker crowd.
Sometime in 1997 or 1998, Shawn was invited to join the private IRC channel called w00w00, which was the main online meeting place for a hacking collective of the same name. Members of w00w00 would go on to have a hand in the formation of dozens of technology companies ranging from WhatsApp to Arbor Networks, but at the time, they were just a bunch of kids trading hacks.23 Under the pseudonymous login handle “napster,” Fanning traded programs and coding advice, trying to impress the other hackers with exploits and programs he scratched together himself.
In the fall of 1998, Shawn enrolled at Boston’s Northeastern University and saw that his new roommates and fellow students were obsessed with finding and trading music files known as MP3s. But finding these files was a complicated process of searching FTP (File Transfer Protocol) sites, Usenet newsgroups and other online repositories. There was also no real way for users to exchange these files easily among themselves. So, late in 1998, Shawn Fanning announced to his fellow hackers on w00w00 that he was working on a program that would make finding and exchanging MP3 files a breeze.
■
FROM THE EARLIEST DAYS, people had dreamed of turning the web into a medium for music. As early as 1993, two students at UC Santa Cruz launched a website called the Internet Underground Music Archive so that artists and musicians could upload and distribute digitized recordings for others to download and listen to. This proved popular, but largely unwieldy for most users, since the size of the music files was too large for the dial-up Internet connections of the day; downloading a single song could take half a day to complete. This changed in the mid-nineties, when a new type of music file was introduced. ISO-MPEG Audio Layer-3, or MP3, was developed at the Fraunhofer Society for the Advancement of Applied Research in Germany and used audio and file compression to create music files that were much smaller in size, but without sacrificing too much in the way of sound quality.
It turns out that the human auditory system is not an instrument that scoops up all the frequencies in a given environment, like a microphone does. What we “hear” is not an accurate representation of reality, but only those sounds that the brain, over the course of millenia of evolution, has determined to be the “most important” sounds. By stripping out the unnecessary (because they were unheard) noises in a sound file, music files could be made much smaller. Most music was easily compressed and a listener was none the wiser. “That’s an undergraduate project,” says Karlheinz Brandenburg, the Fraunhofer researcher who is called the “father” of the MP3.24 But the human voice was far trickier. It turned out that the key to mastering the nuances of human singing was an obscure a cappella recording of a minor hit from the 1980s, Suzanne Vega’s “Tom’s Diner.” Brandenburg successfully tweaked the MP3’s compression algorithm by listening to “I am sitting / In the morning / At the diner / On the corner . . .” over and over again, maybe 10,000 times, before he got it right. “To get it to the level that it’s really perfect, or near-perfect, for everything,” says Brandenburg, “that was work.”25
The resulting files were small enough to be useful in a low-bandwidth era, but MP3 technology further benefited from another technological leap that was occurring at just the same time: computer storage was exploding. The web had been born in an era when the average computer hard drive was still measured in megabytes. The first gigabyte hard drives only became commercially available in the mid-1990s,26 and by 1999, CNN was trumpeting the arrival of 5GB, even 10GB, hard drives.27 That amount of storage might seem woefully small for even a smartphone these days, but in the late 1990s, it was a massive amount, more than enough to store not just numerous songs, but entire albums worth of MP3s.
The media was there, the storage was there, and just as serendipitously, the ability to play this media arrived on the scene as well. In 1997, a nineteen-year-old college dropout named Justin Frankel released a software program called Winamp, which allowed users to easily organize and play MP3s on computers. Winamp was downloaded by more than 25 million eager MP3 devotees, and Nullsoft, Winamp’s parent company (which Frankel had formed with the Internet Underground Music Archive’s Rob Lord), was sold to AOL in 1999 for around $100 million.28
In a way, Shawn Fanning was trying to solve the final piece of this puzzle: a search engine for MP3s. But since most MP3s were sitting on individual users’ computers, he needed to find a way to search other people’s hard drives, not public webpages. That way, if you wanted to find a particular song, you could simply figure out who had it on their computer and get it directly from them. You would share the songs on your hard drive as well, thereby keeping the karmic cycle going. Fanning’s MP3 search program would be networking in its purest form; it would be a literal peer-to-peer exchange.
“It felt like this way of sharing media between people could be used for sharing anything,” Fanning would say later. “It also felt like this whole model for sharing media was superior to, like, going and buying an album. . . . Basically to have access to the entire universe of recorded music. . . . In every way it seemed like a better system.”29
■
IN A FEW SHORT WEEKS, Fanning coded up a rough version of a program, which he named after his nom-de-hacker, Napster. As was the custom, he turned to the other hackers in w00w00 for tips and advice. Among those other w00w00 users who began chipping in to contribute to the program was a slightly older, slightly more sophisticated coder named Jordan Ritter (w00w00 handle: “nocarrier”) and a less technically savvy but more ambitious w00w00 hanger-on named Sean Parker (nickname: “nob”). Ritter would eventually take over the sophisticated back end of the Napster system, developing the complicated server connections, search algorithms and ne
tworking details that would allow users to search each other’s computers and download MP3s directly among themselves. And as for Parker’s contribution? Well, Sean Parker wanted to turn Napster into a business.
Despite the populist image it cultivated later, Napster was conceived of as a business from day one. The Napster phenomenon was covered in the press as some sort of grass-roots movement that bubbled up out of nowhere, largely because that was the image Napster, the company, later fed to the press. But the truth is that long before Napster was a multimillion-user phenomenon—before Napster even had users numbering in the tens of thousands—the idea was to turn Napster into a billion-dollar company. This inclination was partially a result of the time Napster was born into; 1998 into 1999, when Napster was being developed, was the height of the dot-com mania. But it was also because the brilliance of the Napster idea was immediately obvious to everyone involved: it was an entirely new way to distribute media. Imagine being able to search and instantly find any song in existence. And then imagine the instant gratification of being able to download those songs and play them right away. Oh, and by the way: all those songs were completely, 100% free of charge, because you were getting them, not from a record store, but from some other, unknown Internet user.
How the Internet Happened Page 22