Googled: The End of the World as We Know It

PART THREE Google Versus the Bears

CHAPTER TEN Waking the Government Bear

While a full chorus of incensed media—advertising agencies, publishers, newspapers, television and telephone companies, and tech companies like Microsoft—complained about the growing power of Google, the Bush administration, steadfast in its belief that a free market provides its own regulation, was silent. Stepping into this breach was Brooklyn-born public interest advocate Jeffrey Chester, executive director of the Center for Digital Democracy in Washington, D.C. Chester founded this two-person organization with an annual budget of two hundred thousand dollars in 2001. He has mounted a ferocious campaign to induce the world’s governments to handcuff Google. Its first petition was filed before the FTC in the fall of 2006, prodding them to investigate how online marketing encroaches on privacy. In the spring of 2007, Chester, then sixty-two, began to press for an antitrust investigation of the rapid consolidation of the online advertising sector, and urged the FTC to reject Google’s proposed merger with DoubleClick. He petitioned the European Commission to do the same.

A voracious reader of trade publications, Chester became obsessed by what he saw as the pernicious power of the Internet to compile data on consumers. Chester is difficult to ignore. His Brooklyn-accented voice is loud and piercing. He hounds people. He speaks passionately and rapidly, leaping in midsentence from privacy to monopoly to a conversation he had that morning with an FTC staffer. He wears horn-rims and short-sleeved shirts with the neck open and the pockets bristling with pens. His tiny office on Connecticut Avenue is adorned with movie posters that assail McCarthyism and corporate power. He has little regard for the advertising industry, but knows that if he railed against commercialism and consumerism it would open him up to attack as a left-wing former social worker, which, of course, he is. So he sticks to the privacy issue. “The basic model for interactive advertising,” he said, “combines this very powerful data-collection business designed to know your interests in a daily, updated way that is then utilized to create very powerful multimedia to get you to behave in some fashion, whether it’s buying a product or liking a brand.”

Marc Rotenberg, the executive director of the Electronic Privacy Information Center, rents a single office to Chester’s organization and works just down the hall. He is in nearly all ways Chester’s opposite. He wears charcoal business suits and has degrees in law and computer science; no pens can be found in his shirt pockets. But he and Chester work closely together to advance privacy protection measures. Rotenberg believes the central question should not be, Is Google invading people’s privacy? Rather, it should be, Why does Google need to collect all of this information?

GOOGLE’S SERVERS NOW CONTAIN a tremendous amount of data about its users, and this database grows exponentially as search and a variety of Google services multiply. With the latest techniques to discern what really motivates consumers—often categorized as “behavioral targeting”—companies and advertisers will know even more. Some forms of such targeting are widely seen as helpful, such as when Amazon extrapolates from the browsing and purchase histories of a customer to recommend books. Other forms might be alarming to the lay consumer. New technology will allow cameras built into television set-top boxes to be armed with algorithmic models that read our facial expressions and tell advertisers what we do and don’t like; Nielsen is investing in brain reading—called NeuroFocus—which is meant to take the guessing out of why consumers react to what they see on a screen or read or listen to.

New smart phones collect enormous amounts of data. Mobile telephone companies gather and store digital data on calls made and received and how long each lasted. In addition, the chips in the phone’s GPS track a user’s location, the length of stay, and other mobile users she is in touch with. Tapping this sort of data is known as reality mining, and is a cousin to Brin’s data mining. Although telephone companies don’t share the names of customers, they have begun to sell this data to companies seeking to market products. Phorm, an American company with offices around the world, proposed to go one step further, approaching telephone and broadband Internet service providers with software that tracks each consumer’s online activities, so that a nameless portrait of each consumer can be created. In return for supplying the data, the telephone and cable companies can open a new revenue spigot. By late 2007, Phorm had done three deals in England that yielded data on two-thirds of Britain’s broadband households.

Publicity about Phorm aroused the ire of Tim Berners-Lee, a senior researcher at MIT and the inventor of the World Wide Web. Because Berners-Lee refused to patent his invention, to cash in financially, or to become a talk-show celebrity, his opinion carries heft. In a rare interview with the BBC, Berners-Lee expressed outrage: “I want to know if I look up a whole lot of books about some form of cancer that that’s not going to get to my insurance company and I’m going to find my insurance premium is going to go up by 5 percent because they’ve figured I’m looking at those books.”

The data did not belong to Phorm or the telephone or cable company, he said. “It’s mine—you can’t have it. If you want to use it for something, then you have to negotiate with me. I have to agree.” This seemed to be the view of many European Union officials, for they were gathering evidence to determine whether to impose restrictions on this practice.

Because the financial rewards will be so huge if corporations can capture and use this data, pressure to do so will increase. Describing the effort to track the identity of an online user, Irwin Gotlieb invokes an imaginary user searching for an SUV: “If you’re searching for an SUV and you price out a couple of them and you go to a site that requires you to register, I now have your name. If you want pricing, dealer costs, you’ve got to give me your name, your e-mail address. Now as soon as you do that, we’ve got more on you. There are lots of ways we can track you. If I’ve identified you with your address, I can go to DMV records and see what cars you own. I can then go to Experian and see what your credit history is. And if I find out you’re in the last month of a thirty-six-month lease on a Land Rover ...” He doesn’t finish the sentence but smiles, as if he’s trapped his prey. “I can’t do that today. In a few years I can.”

There’s no question that new technologies spur ways to improve services, and also to quietly redefine privacy. A home becomes less of a castle. Google’s Street View has cameras on city streets that reveal street activity and traffic—and also license plate numbers and the faces of a passersby. If you know the address of a celebrity (or an old girlfriend), it allows close-ups. Similar closed-circuit television cameras on streets and in stores help police solve crimes, but might also catch an old paramour or celebrity smooching. Cell phones help parents track the location of their children, but can track much more. Part of YouTube’s appeal is that it shares private moments. Facebook is a glass house, a complaint sometimes voiced by parents of teenagers who share their sexual exploits. Marketing companies create ads that annoyingly pop up when people are online, offering a premium—free telephone calls—in exchange for permission to monitor your activities.

Google’s Web site acknowledges that it collects information about its users, but not names or other personally identifying information. It does, however, collect the names, credit card information, telephone number, and purchasing and credit history of those who sign up for such features as Google Checkout, a service that enables customers to make online purchases. In its five-page Checkout privacy policy, Google said that it may also “obtain information about you from third parties to verify the information you provide.” Google said it “will not sell or rent your personal information to companies or individuals outside Google”—unless individuals give their “opt in consent.” But even if consumers choose not to opt in, Google still retains a wealth of personal information, which it is free to use to better target search or YouTube advertising. This data, mingled with its search data and the data gathered by DoubleClick, induces privacy anxiety. Companies like eBay and Amazon.com, among others, also retain credit card and other personal information, but there is a difference: Google uses this data to assist advertisers, and eBay and Amazon do not have advertisers to assist.

This is where the specter of Big Brother enters the discussion. “In 1984, Winston Smith knew where the telescreen was,” observed Lawrence Lessig. “In the Internet, you have no idea who is being watched by whom. In a world where everything is surveilled, how to protect privacy?”

Big Brother comes in different guises. Under the federal Patriot Act rushed to passage after 9/11, the executive branch can, without a warrant or users’ knowledge, gain access to what Americans e-mail, search, read, say on the telephone, watch on YouTube, network with on Facebook, or purchase online. There are tens of millions of surveillance cameras on our streets and in buildings. Candidates for public office can be harmed by damaging leaks to reporters of the most private information, as nearly happened to Supreme Court nominee Clarence Thomas when his in-store video rentals were leaked, suggesting he was partial to pornographic films. Advertisers can pay Google and other companies to access better targeting data. Telephone or cable companies can offer free services in exchange for a record of a consumer’s every click, encoding the knowledge of every song they listen to, product they buy, ad they like. In recent years, AOL (among other companies) was publicly embarrassed when it extracted and shared the names of some users from the Internet protocol address found on every browser on every computer.

Google may be viewed with suspicion by many media industries, but it enjoys a well-deserved reputation for earning the trust of users. In its 2007 annual ranking of the world’s most powerful brands, the Financial Times and the consulting firm Millward Brown awarded Google a number one rating. Still it is hard to imagine an issue that could imperil the trust Google has achieved as quickly as could privacy. One Google executive whispers, “Privacy is an atomic bomb. Our success is based on trust.” The Interactive Advertising Bureau’s Randall Rothenberg employs a different image: “Privacy is one of those third rails with unknown amounts of electricity in it. People think about advertising and worry. That’s part of America’s don’t tread on me attitude.” Yet Rothenberg is unsure about the amount of electricity in the privacy issue because he is acutely aware that Americans hold contradictory attitudes toward privacy; they tend to distrust governments, businesses, or advertisers that can spy on them, yet “if they really cared enough about privacy they’d never have credit cards. They’d never subscribe to cable TV” They wouldn’t parade their most private thoughts on Facebook. So how does a company like Google locate and not step on the third rail?

To find that third rail requires a degree of sensitivity not always found in engineers. Yet Google has at times been sensitive. When, in 2006, the Justice Department subpoenaed a number of Internet companies to turn over all child pornography search queries, Google, opposing the request as a “fishing expedition,” took the U.S. government to court and won. (The judge ruled that Google had to turn over fifty thousand suspicious URLs, but none of its user’s names.) “Unfortunately,” said Schmidt, “our competitors”—including Time Warner, Yahoo, and Microsoft—“did not hold the same position, and complied.” He is outspokenly critical of the Patriot Act, which he believes violates privacy and grants the president too much power. Schmidt implicitly agrees that privacy is a do-not-step-on third rail: “If we violate the privacy of our users, then we’ll be hosed.”

Trust is essential to Google, as it is to much of modern commerce. If we didn’t trust waiters or Amazon, we couldn’t use credit cards. If banks didn’t trust that we’d pay back loans, they would not grant mortgages. Sergey Brin notes that users don’t like intrusive ads or junk mail, and Google deserves credit for siding with users against these, even sacrificing revenues to do so. From a user’s perspective, Google is not wrong when it says the more information it has about our search histories, the better the search results will be because it can anticipate a user’s intent. And it is also undeniably true that many users search the ads to comparison shop. “We think of the ad as content,” said Google’s vice president of engineering, Jeff Huber. The portrayal of Google as Big Brother frustrates people who work there. “It’s a fear of the possibility rather than the reality,” said Huber.

FEAR OF “THE POSSIBILITY” was enough to motivate privacy advocates. In May 2007, the Bush administration decided that the Federal Trade Commission (FTC), not the Justice Department, would look at whether the merger might violate antitrust laws, and the commission quietly began a preliminary investigation. Chester pushed for the FTC to broaden its probe to include privacy, and the agency scheduled a town hall meeting in November to explore those issues. “I got the hearing,” Chester boasted. “There’s nobody watching the store.”

Chester is not the stereotypical lobbyist, the ingratiator, the affable fund-raiser, the guy you’d join for a drink or a steak dinner at the Palm. Chester annoys people, concedes a senior FTC staffer, “but he is responsible for the hearing.” Chester helped persuade Democratic senators on the Judiciary Committee to conduct a one-day hearing in September 2007 focused on the merger and online advertising. Knowing that European governments have more stringent privacy protections, he urged the EU to hold up the merger until Google better explained its privacy policies. Chester stayed in touch with government regulators elsewhere. He filed additional complaints with the FTC. He had no expectation that the FTC would reject the merger. His best hope was that the FTC would compel Google to admit that its warehouse of data required more stringent privacy safeguards. Chester and Rotenberg were more hopeful that the European Union would reject the merger; and privately, Google officials conceded that an EU rejection would scuttle the deal.

By the end of 2007, it was apparent that privacy issues were gaining traction, spurred by a crescendo of news stories about real or potential invasions of privacy. Many were chilled when President Bush declared that under the Patriot Act he did not need judicial or congressional approval to wiretap the phones of anyone the executive branch suspected of consorting with, or knowing, alleged terrorists. It was revealed that telephone companies, at the request of the Bush administration but without court approval, turned over oceans of data concerning the phone calls and e-mails of individuals.

There is often an inherent conflict between privacy and Google’s belief that data is virtuous. Eric Schmidt, as we’ve seen, said that the more Google knows about a user, the better the search results. His and the founders’ ideal? For Google to know enough to be able to anticipate the user’s true intent in a search query, eventually getting to the point where Google could improve the user’s experience by supplying a single answer to a question. Better targeted ads, Google believes, serve the consumer as well as the advertiser. In an October conference call with analysts to discuss Google’s third quarter and its new products, Schmidt said, “We’re working on expanding our breadth of ads, offering all sorts of new types of ads—gadget ads, video ads, others coming. And each of these initiatives gives advertisers new and interesting ways to build relationships with their customers. So by building these deeper ad solutions, we really can deliver more value.” More “types of ads” equals more data, which equals more assistance to advertisers, which can be a service to consumers, but also an invitation to nibble at privacy constraints. Schmidt insists Google would never risk violating user privacy because Google’s success pivots on user trust. Rotenberg counters, “If people knew what Google was doing, they’d lose trust.”

In a voice as steady as a dial tone, Schmidt said Google users can click to “opt out” of allowing Google to track its cookies by clicking on the Privacy choice on the Google home page and following the instructions, a feature most users are probably not aware of or have a difficult time finding. Rotenberg, Chester, and others say that instead, Google should allow users to opt in, meaning that they would have to actively volunteer their cookies. That would be a mistake, Schmidt said, because the quality of a Google search would be inferior if it stored none of the cookie data; that’s what helps Google better answer a search question. Without information about the user, Google would not be able to narrow search results based on prior searches. Without a cookie, an American living in Paris would receive search results in French, not English. Schmidt disputes the notion that advertisers have become as important to Google as the user, reciting the company’s official first two principles: One is “the quality of the search as seen by the end user. The second one is the quality of the ads as seen by the end users, not by the advertisers.”

When asked why consumers should trust that Google would not abuse the private data it collects, Sergey Brin in 2007 told me that the fears people have are tied to a distaste for advertising and to a fear of Big Brother, which is sometimes “irrational.” He wondered: “How many people yesterday do you think had embarrassing information about them exposed as the result of some cookie? Zero. It never happens. Yet I’m sure thousands of people had their mail stolen yesterday ... I do think it boils down to irrational fears that all of a sudden we’d do evil things.”

Irrational or not, Google was assailed from many sides and compelled to play an unaccustomed role: defense. Nick Grouf, CEO of Spot Runner, an advertising/marketing agency that counted among its investors Martin Sorrell’s WPP Group, believed Google was engaged in too many battles. He said that traditional media woke up to the Google threat not when the company did its IPO or was sued by book publishers, but when it bought YouTube. “When you pay $1.6 billion for a site that is on the cover of every newspaper and magazine, and is the centerpiece of the zeitgeist as the future of media”—suddenly Google was widely perceived as a media company. By 2006 and into 2007, Grouf said, Google was battling with television and newspapers and book publishers and Microsoft and eBay and advertising agencies. “It’s hard to compete on all fronts. And people start to whisper: ‘These guys have gargantuan ambitions.’”

IN NOVEMBER, the FTC held a two-day town hall meeting on privacy, a series of tame panel discussions that became more a seminar than an inquisition, disappointing Jeff Chester. The commission decided that the often-baffling issue of privacy would be excluded from the decision of whether to approve Google’s acquisition of DoubleClick. The focus, instead, was on whether the marriage was anticompetitive. It was difficult to argue that the merger harmed competition when, within months, companies such as Microsoft and Yahoo and AOL and the WPP all acquired digital advertising companies of their own. The FTC prefers “to wait for a violation before we act,” an agency official said on the eve of the approval of the merger. The EU did compel Google to make concessions and to tighten its privacy policies, but it, too, would approve the merger.

By mid 2007, Google was worried about the many restive bears it had provoked. It began to reach out to Washington. To allay privacy concerns, the company announced that it would reduce from two years to eighteen months the information it keeps in its database about the Web search histories of its users. Claiming that privacy laws were out of date, Google put out a press release proposing uniform international privacy rules and perhaps laws that recognized how the Internet and technology posed new privacy challenges. Instead of one “uber cookie” that permanently tracks a user, Google said it was experimenting with “crumbled cookies” that would disappear over time.

The public battles probably made Google’s executives somewhat wiser. Google was only guilty, they believed, of naivete, not arrogance. “The product brand was very strong,” said Alan Davidson, Google’s senior director of government relations and public policy, who is a computer scientist as well as a lawyer and who oversees Google’s Washington office. “The political brand was very weak. Because we were not here to define it, it was being defined by our enemies.” He paused a moment, and added, “Enemy is a strong word. It was being defined by our competitors.” Gigi Sohn, the president and cofounder of Public Knowledge, a nonprofit organization that lobbies for both an open Internet and more balanced copyright laws, said that like many Silicon Valley companies, Google chose to have a smaller presence in the nation’s capital. But Google was more extreme, she said. “They were almost alone among Silicon Valley companies in failing to recognize that you have to play in the sandbox. If you want progressive spectrum policies, the free market does not ensure that.”

Google’s one-man operation in Washington expanded in 2007 to include twenty-two staffers. Among them were Jane Horvath, a former senior privacy attorney in the Bush administration’s Justice Department; Johanna Shelton, former senior counsel to Democrat John Dingell, then chairman of the House Energy and Commerce Committee; Robert Boorstin, a former speechwriter for President Bill Clinton; and Pablo Chavez, former chief counsel to Republican senator John McCain. To advance its Washington agenda, Google had established its own PAC (NETPAC), and soon hired three outside firms to lobby on its behalf: the mostly Democratic Podesta Group; King & Spalding, where Google relied on former Republican senators Connie Mack and Dan Coats; and Brownstein Hyatt Farber Schreck, which had recently hired Makan Delrahim, a former deputy assistant attorney general who’d been in charge of the Antitrust Division in the administration of George W Bush. “We’ve been under the radar, if you will, with government and certain industries,” observed David Drummond, the Google senior vice president who oversees all of the company’s legal affairs and policy interaction with governments. “As we’ve grown, we’re engaging a lot more.”

The most immediate concerns of Google’s Washington office were the privacy issues raised by the acquisition of DoubleClick. By the end of 2007, Google was battling the image that it was the Microsoft of 2000. “No question that people here regularly discuss Microsoft’s experience and use that as a cautionary tale,” said Elliot Schrage, the vice president of global communications and public affairs. On the subject of Microsoft, Brin said, “Microsoft is a bit of an unusual company. They don’t seem to like any of us being successful in the technology space.”

So Google sought to demonstrate that it was reaching out to media companies as well as to Washington. To preserve copyrights, YouTube announced that it was testing new antipiracy software to block unauthorized content from being uploaded and viewed. In an ecumenical spirit, the word partnership was constantly invoked by Google executives. Repeatedly, they celebrated its “more than 100,000 partners,” the more than three billion dollars it then distributed annually to Web sites and mostly small business partners in its AdSense program. As 2007 progressed, said General Electric’s Beth Comstock, relations with Google thawed and by summer, G.E.’s NBC negotiated for Google to sell some of the network’s remnant ads. “In the end, I was less concerned that Google was out to replace our entire sales force,” she said.

Google, however, was still clear-eyed about the inevitable gap between its engineers and traditional media. Google’s engineering prowess would, inevitably, make the consumption of media and the selling of advertising more efficient, Larry Page told me one afternoon as we sat in the small, bare-walled conference room steps from his office. So was it inevitable, I asked, that “Google would sometimes bump into traditional media?”

Without hesitation, he corrected me. “I would say, always,” he said in his deep baritone, emitting a subdued chuckle. His was not a boast; rather, it was a candid recognition of reality. He believes Google’s engineers can eradicate most inefficiencies if given the time and resources.

Page had reasons to feel confident. Google had a great year in 2007. Measured by growth, it was Google’s best year, with revenues soaring 60 percent to $16.6 billion, with international revenues contributing nearly half the total, and with profits climbing to $4.2 billion. Google ended the year with 16,805 full-time employees, offices in twenty countries, and the search engine available in 117 languages. And the year had been a personally happy one for Page and Brin. Page married Lucy Southworth, a former model who earned her Ph.D. in bioinformatics in January 2009 from Stanford; they married seven months after Brin wed Anne Wojcicki.

But Sheryl Sandberg was worried. She had held a ranking job in the Clinton administration before, joining Google in 2001, where she supervised all online sales for AdWords and AdSense, and was regularly hailed by Fortune magazine as one of the fifty most powerful female executives in America. Sandberg came to believe Google’s vice was the flip side of its virtue. “We’re an engineering company in that products come first,” she said. “A lot of the reason we’re winning is because our engineering is better.” Reminded that she once was quoted as saying Google made a mistake in not speaking to publishers and answering their questions before announcing plans to digitize all books, she added, “Sometimes we make mistakes here because we move too quickly”

Eric Schmidt would, inadvertently, prove her point. In August 2007, he piloted his Gulfstream G550 to Aspen, Colorado, to give the keynote speech at a dinner held by the free-market oriented Progress and Freedom Foundation. In the speech, he described four “basic principles,” as he referred to them, that he believes are vital for media and tech companies to embrace: freedom of speech, universal broadband access, net neutrality, and transparency. Missing from his prepared remarks were thoughts about privacy and copyright—and how far Google might push the permissible boundaries. (When, for instance, does anticipating a user’s wants become an intrusion? When does fair use become copyright infringement?)

A few weeks later, seated in his tiny conference room on the Mountain View campus, I discussed that speech with Schmidt. Why, I asked, didn’t he mention privacy in his Aspen talk?

There was a long pause before he said, “No particular reason. It’s sort of a given. If we violate the privacy of our users, they’ll leave us.”

And why no mention of copyright?

“Maybe it was the altitude! I was just chatting away.” Besides, he said, copyright “was not an absolute right” and had to be balanced by fair use.

Isn’t it true that Google wants to push the envelope on privacy and copyright?

“That’s probably correct,” Schmidt conceded. “If there’s a legal case, we’re going to favor the legal one that favors users.”

“Google, if it were a person, has all the flaws and all of the virtues of a classic Silicon Valley geek,” said Columbia’s Tim Wu, who between jobs teaching law worked for a spell in the Valley. “In some ways, they are very principled.” He cited Google’s 20 percent time, saying that few “money-crazed companies would allow” such a thing. “But they have this total deaf ear to certain types of issues. One of them is privacy.” Why? Because, he said, “They just love that data because they can do neat things with it.”