by Steve Levy
The product gained a following at Google, but every time Horowitz took it to a GPS, the founders would pounce on it. Brin wanted more from it. Taco Town became a priority for Brin, and in mid-2009 he actually moved his office to the apps group so he could monitor it more closely.
It was an indication of Google’s confused strategy that Taco Town’s development proceeded even as the company announced Wave with fanfare and hosannas. When pressed, Horowitz would concede that Taco Town’s functions overlapped with those of Wave. “In the worst-case scenario, Wave is a concept car,” he would say. “General Motors doesn’t build every concept car.”
As the team prepared for Taco Town’s rollout in early 2010, the product added more features, many of which duplicated Facebook functions. It also added location information to the Tacos. But the minipostings would no longer be called Tacos; Google renamed the product Buzz to reflect the crackling interaction it would presumably generate. The moniker more accurately reflected the product’s purpose but lacked the irreverent pizzazz of the original. Nonetheless, excitement ran high at Google that the company had finally cracked the social problem. Thousands of Google employees used Buzz in the dogfooding process.
The evening before the launch the team gathered for a rehearsal. Then the PR people joined them. Vic Gundotra, Google’s most polished presenter, gave a brief demo of Buzz’s mobile abilities. Horowitz delivered a product overview. Gmail’s product manager went into details. (Sergey Brin wasn’t at the rehearsal but would attend the launch the next day.) Then, in a dry run of the postpresentation Q and A, Google’s PR staff pretended to be reporters asking their toughest questions. Why isn’t Facebook in there? What about Wave? None of the questions touched on whether there might be privacy concerns in building an instant social network based on one’s email contacts.
Indeed, on launch day—February 9, 2010—none of the reporters in attendance asked probing questions about the new product’s privacy settings, and the first wave of articles about the product was generally positive. A number of the Google executives at the event, including Horowitz and Brin, left Mountain View soon afterward to attend the annual TED Conference in Southern California. But within forty-eight hours, Buzz ignited a privacy crisis as intense as the Gmail privacy conflagration in 2004.
The problem lay in a feature that Google was most proud of. Previously, new users of social networking services had been confronted with the annoying chore of gathering friends and contacts to construct their cohort. Google felt it had solved this problem with Buzz. When a Gmail user clicked the single button that signed him or her up for Buzz, a social network instantly appeared, based on one’s email contacts. When this feature was tested internally, the employees trying it out loved it.
But when the general public tried Buzz, some users discovered unwanted—even horrendous—consequences. By looking at a Buzz user’s profile, other Buzz users could see that person’s social network. Since the network had not been carefully built contact by contact, it was entirely possible that it might include a connection that a user might not want exposed to a larger audience. (Certain contacts could indicate that someone was seeking alternative employment or spilling secrets to a reporter.) As described in a February 10 posting in Business Insider, “The problem is that—by default—the people you follow and the people who follow you are made public to anyone who looks at your profile … someone could go in your profile and see the people you email with and chat with most.” The settings for what was exposed to the public, as well as which contacts were included in one’s network, could address this, but most users follow the standard settings.
Google had made a critical error. Its employees differed from the general population. For one thing, their email contacts are largely with other Googlers. So few of them were concerned that the networks instantly constructed by Buzz drew on their Gmail contacts. Instead, they were motivated to explore its features and find those that worked for them. (Brin boasted to The New York Times that he had used Buzz input to write his op-ed defending the book settlement.) As a result, the product team—as well as the usually vigilant Google privacy squad—missed something that became obvious as soon as the product was released to a population whose electronic correspondence often held secrets. Nicole Wong, the lawyer in charge of Google’s policy operations, later admitted the mistake. “The on-boarding [dogfood] process is not like doing it in the wild, and the social network of 20,000 Googlers is not like being on the Internet. That process failed us.”
The outcry was instant and loud. A domestic violence victim complained that Buzz had exposed her blog comments and reading habits to her abusive former spouse, revealing information that hinted at her whereabouts. Foreign Policy’s Evgeny Morozov suggested in a blog post that Iranian and Chinese government goon squads might instantly check Buzz accounts of dissidents to analyze their connections. Even Google’s former policy head Andrew McLaughlin wrote—in a Buzz post!—that “Google exposes the people you email most by default, to the world. This violates my sense of expectations.” Privacy activists prepared formal complaints to the Federal Trade Commission. The privacy commissioner of Canada, in a letter cosigned by data protection officials of nine other nations, charged Buzz with “a disappointing disregard for fundamental privacy norms and laws.”
Google quickly set up a war room, populating it with not only policy and PR people but engineers working to alter the product. Still at TED, Bradley Horowitz felt blindsided. “We knew we were doing something dangerous, taking a private space and opening it up to a social activity,” he said. “But we thought that after Facebook and other services this was something people were used to.” He felt, however, that the storm, while intense, would soon pass. “We’ll get through this,” he promised. Indeed, in record time, the engineers made changes in the product. They changed the default settings to allow people to more easily keep contacts private and block unwanted followers. Eventually, Google would address virtually all the complaints of privacy advocates.
But a couple of months later, Horowitz admitted that the damage to Buzz had been deep. “We should have known people were gunning for us,” he said. The privacy flap, he admitted, was “a scar that will stay with the product forever.”
As Buzz stumbled into its first summer, the product looked like a goner. Meanwhile, Google quietly announced an end to Wave. Though its 2009 demo had arrived like a killer swell, by the time it reached the shore, it couldn’t support a boogie board. “Wave has not seen the user adoption we would have liked,” wrote Urs Hölzle in his August 4, 2010, blog post announcing the termination. The move was little noticed, because Wave had so little adoption. Two months later, the head of the Wave team, a star engineer named Lars Rasmussen, announced that he was leaving Google to join Facebook.
Google still hadn’t cracked social. But that didn’t mean it would stop trying. “If we see a way to deliver a benefit, should we simply not go there because there’s another company there?” asked Nicole Wong, shrugging off the privacy blunders of Buzz. “If Facebook were your only option, would that end up being a good thing?”
Facebook wasn’t the only new competitive challenge Google faced. Its failure to prevent the merger of the search services of its next biggest rivals, Microsoft and Yahoo, had allowed those two companies to merge their user base, with Microsoft providing the search technology. After many years of relatively poor efforts, Microsoft was now committed to spending hundreds of millions of dollars to build a competitive engine. To head the team, it hired the scientist Qi Lu, a forty-eight-year-old whose tireless work habits were legendary. Those regarding this as a coup included Google’s search czar, Udi Manber: “I have the highest regard for him,” he said. Microsoft called its new search engine Bing, and it was launched in June 2009 by CEO Steve Ballmer with great fanfare.
In terms of search quality, Bing did not intimidate Google. Its relevance algorithms were basically no different from those in the previous version of Microsoft’s search, much less likely to
draw out the Audrey Fino–like needles in the Internet haystack. Eventually that could change, as Microsoft would supply Bing to Yahoo for the latter company’s search engine. That would provide Microsoft with a critical mass of users to run the thousands of constant experiments necessary to improve search quality. “The algorithm is extremely important in search,” said Microsoft’s VP of core search, Brian MacDonald. “But it’s not the only thing.” He compared it to a car: the engine is very important, but there are all sorts of other reasons to choose a given model. MacDonald said that Google, with its dependable ten blue links, “still looks like your father’s Oldsmobile. If you were Rip Van Winkle and went to sleep twelve years ago and woke up today, you’d still have no problem using Google.”
Though this wasn’t really true—Google had previously spiced up its blue links with “one boxes” for things such as weather, travel, news, and video—Bing did look flashier than its entrenched competitor. That was most striking in video search: Bing presented search results in an array of thumbnail depictions of the most relevant videos, offering instant playback. Also, Microsoft had tried to identify weaknesses in Google search, purchasing innovative companies that specialized in those areas.
Publicly Google presented an attitude of calm engagement to the public, with Brin saying to reporters that his company welcomed the enhanced competition. But in Building 43, there was something of a freak-out. The search team set up a war room, hurriedly launching an effort dubbed the skunkworks. (That appellation, first used at Lockheed aircraft during World War II, is a generic term for an off-the-books engineering effort that operates outside a company’s stifling bureaucracy. The fact that Google needed a skunkworks was telling in itself.) Its OKR was to change the look of search 25 percent within a hundred days. Within the search team itself, Googlers engaged in finger-pointing and recriminations. Months earlier, Google search engineers had presented their bosses with a project that streamlined video search results and offered instant playback—but Google had rejected it. Now the search interface team was more open to change. Very quickly, Google instituted a couple of distinctive visual changes to its home page. In one, the search box was “supersized,” made about a third bigger. The text size of the search queries users typed in was similarly boosted. It stood as sort of a symbol that Google was still the search company. Some users were startled by the change. “People were saying that the search box was so big that it could actually eat you whole,” Marissa Mayer later said. But it worked—as Mayer explained, Google ran later A/B experiments that restored the box to its original size. Hundreds of people wrote emails complaining. “They said, ‘What’s going on with the search box? It’s so small there’s not even room to type!’” In another refinement, Google simplified the initial view of the home page by removing everything except its logo and the search box; when the user moved the mouse or typed, then the rest of the text would come into view.
Though the skunkworks began with a sense of urgency, the pressure eventually subsided as it became clear that the survival of Google didn’t hinge on its efforts. At one point, Larry Page bounced its efforts, complaining that the redesign looked too much like Bing. Eventually, Google did release a revamped search results page, using a three-column view: in addition to the organic search results and the ads, there was a column to the left with various search options. But it wasn’t a dramatic shift. Nor did Google need one. By then several months had passed, and Bing’s gains in market share were minimal.
The Bing challenge was a healthy prod for Google. It energized the search team and forced a rethinking of how Google did its interface. When Google’s executives met in 2010, the main topic of discussion was not search but Mark Zuckerberg.
That March, Urs Hölzle sounded an alarum that evoked Bill Gates’s legendary 1995 “Internet Sea Change” missive to his minions at Microsoft. Just as the Internet threatened Microsoft back then, in 2010 the sea change to a more people-oriented Internet—social media—was becoming a problem for Google. Hölzle said that the challenge required a decisive and substantial response, involving a significant deployment of personnel—right away. The memo became known as the Urs-Quake.
At the time Google had just completed a renovation of 2000 Charleston Road, only a few hundred yards from the main Googleplex headquarters in a complex of four-story structures once owned by the Alza drug company. (The Chrome team was next door, in Building 1950.) After the Urs-quake, the top two floors of 2000 became the nerve center for Google’s social network. Vic Gundotra led the team, joined by Bradley Horowitz. Teams of engineers from various outposts of the company moved into the building, and almost on a daily basis Google’s top executives would cross Permanente Creek to strategize.
The project’s internal code name was Emerald Sea. When Horowitz typed those words into Google Image Search that spring, the top result was an 1878 painting by German immigrant artist Albert Bierstadt. It depicted a tumescent seascape, dominated by a wall of surf that had already upturned a pitiful sailing ship. Horowitz commissioned a pair of art students to copy it onto the wall facing the fourth floor elevators in Building 2000. It was the perfect illustration of the Google mind as it approached the project. “We needed a code name that captured the fact that either there was a great opportunity to sail to new horizons and new things, or that we were going to drown by this wave,” Gundotra would explain.
Gundotra rejected the perception that Google’s DNA, rooted in the primacy of algorithms, was unsuited to accommodate the social networking revolution. To the contrary, he felt Google had unique assets that could help it take the initiative in the field, if only it would atone for its “past sins” of snubbing a social approach. He outlined an ambitious plan that would involve a people-oriented remaking of almost every aspect of Google, from YouTube to search.
Oh, and Google would launch this effort in a hundred days.
Horowitz later described this as “a wild-ass crazy, get-to-the-moon” goal, setting an impossible standard to underline the importance of the effort. A project such as Emerald Sea—which came to include eighteen current Google products, with almost thirty teams working in concert—was complicated and challenging, with milestones more appropriately measured in months, not days. Indeed, on the hundredth day after that May meeting, some time in August, Emerald Sea was nowhere completed. But its leaders were satisfied with the working prototype and had given it a new name: “+1.” This was the term that Googlers and other geeky types would use to respond to an enticing invitation. If someone said he was headed to see the Tron sequel, you’d respond, “Plus one!”
The long delay took its toll. During the months that Google worked on Emerald Sea, Facebook became bigger and scarier. It also poached more of Google’s talent. Then Mark Zuckerberg was named Time magazine’s Man of the Year, and Facebook’s estimated market value reached $50 billion. In Silicon Valley, people assumed that delays in Google’s “Facebook killer” hinted at another failed effort in social networking, a harbinger perhaps of a fall from primacy for Google itself.
Still, Gundotra and Horowitz were energized by what they felt were significant innovations in the initiative, and believed that Emerald Sea would finally establish itself as a primary player in the crucial area of social software. “This is the next generation of Google—it’s Google plus one,” said Gundotra.
In its forays into other areas, like phones, videos, maps, applications, and operating systems, Google had not acted in response to competition. If it had a good idea, it simply pursued it, no matter who was occupying the space. This project was more strategic, even conventional. “It’s a good thing that Google is putting its weight behind social networking, but it’s reactive self-interest, not from a place of idealism,” said one key team member. “It’s not Google at its best, which is truly, truly pioneering. Whereas this thing is clearly more of a reaction to Facebook.”
Twelve years after Larry Page and Sergey Brin decided to go all in with a company they called Google, their empire was broad, the
ir influence massive. Google’s revenues were now approaching $28 billion on an annual rate. (Facebook was taking in no more than a billion dollars a year.) What’s more, even once-skeptical analysts were conceding that YouTube was about to turn a profit. Defying expectation, Google’s Android mobile operating system was thriving: every day more than 200,000 users activated phones running the Android OS. (Eric Schmidt was giddily proclaiming that Google would have no problem eking out $10 in revenue from each user, adding more billions to Google’s bottom line.) And certainly Facebook had no answer to Google’s infrastructure of data centers, its comprehensive collection of global maps and imagery, or its giant learning brain that confounded expectations of digital performance in language comprehension, translation, and voice recognition.
Yet Google felt under siege. Some policy people at Google—now numbering hundreds of lawyers, privacy specialists, and PR experts—called 2010 “the summer of war.” Eric Schmidt was getting flak for a remark he had made about privacy to the effect that young people should be offered a onetime opportunity to assume a new identity so as to distance themselves from embarrassing activities stored in Google’s indexes. (“He was making a joke!” howled Google’s PR people in vain. Maybe so, but the wrong person was making the joke.) Investigators were zeroing in on Google’s Street View Wi-Fi grab. The halo effect of leaving China had worn off, with critics hinting that Google had been self-aggrandizing and naïve in its abrupt decision. And on August 9, Google startled even its most ardent supporters with an unusual announcement.