Dogfight: How Apple and Google Went to War and Started a Revolution (12 page)

For Gundotra, the list of things that made the iPhone revolutionary was endless: The iPhone was beautiful. Apple was free to control it without dreaded carrier intervention. It was the first device powerful enough to run Google’s applications the same way they ran on a desktop. And it had a full Internet browser that allowed Google’s search ads to appear and work normally. This was great for Google because it would help its applications and search ads become even more ubiquitous. It was also great for Google because, as Gundotra had predicted, it was terrible for Microsoft. Microsoft’s power stemmed from its Windows and Office monopoly on desktop and laptop computers. It had little power on mobile phones. Despite Schmidt’s fears about Windows’ traction on mobile phones, Gundotra believed the iPhone was such a leap forward, Windows’ progress would come to a crashing halt. He thought the initial high price of the iPhone was a red herring. Apple would drop the price if consumers resisted.

All of this seemed obvious to Gundotra in the fall of 2007, but it was not to many other Googlers. “People thought it was crazy,” Gundotra said. “Smartphones were a tiny percentage of the mobile business back then [2 percent], so I was accused of believing in the Apple hype. ‘If you think people in India and China are ever going to be able to afford a phone that is seven hundred dollars [$499 for the cheapest model], you’re smoking dope,’ they said.” While he had the support of Schmidt, Brin, and Page, many thought he was challenging a foundational tenet of Google’s culture—that it was a company that played nice with everyone. Google’s success on desktop computers depended on its getting search and other applications to run on
all
software platforms—OS X, Windows, Linux—and all Internet browsers. Getting behind one partner to the exclusion of others was not the way to do that. “I softened the blow a little bit by saying we were shutting down development on all but five smartphones. But it was a very controversial decision. Culturally at Google it was unthinkable that you would not build for every BlackBerry or Windows Mobile phone. Engineers in European Google offices were very angry we wouldn’t support various Nokia phones. People had a hard time seeing [that smartphones would become so important], especially the iPhone. People on my staff quit. It was brutal. While [Microsoft CEO Steve] Ballmer was famous for saying that [the iPhone would fail] publicly,
everyone
believed that. They just thought it was going to go away.”

*   *   *

The Android team was particularly troubled by Gundotra’s emergence. Its members had not only kept a low profile at Google since the company’s purchase in 2005, they’d successfully kept most of Google from even knowing about their project. Now, with Gundotra pushing Google’s mobile agenda much harder, with the iPhone actually available for sale and with early prototypes of their own touchscreen phone—the Dream—now visible in the office, they were going to have to acknowledge and defend what they were doing long before they felt ready to do so. If forced into choosing between Gundotra’s iPhone apps and Android in 2007, it seemed obvious that Schmidt, Brin, and Page would choose the iPhone. The Android was more than a year from even being a product. “That’s when it all [the tension between the two projects] became real,” a former member of the Android team told me. “That’s when they [Android] started to test-drive phones and talk to T-Mobile about how much they were going to spend on marketing. That’s when you started to see that this thing [Android] was going to get bigger and bigger as it went along.”

Up until then Android had been like Google’s mistress—lavished with attention and gifts but still hidden away. This secrecy wasn’t Schmidt’s, Page’s, or Brin’s idea. It was Andy Rubin’s. Rubin didn’t want anyone to know about his project. Like most entrepreneurs, he’s a control freak, and he believed that the only way he could succeed with Android was to run the operation as a stealth start-up inside Google. Google was only nine years old then, but for Rubin, the company was already too slow and bureaucratic. Ethan Beard says he remembers that the (non-Android) part of Google had just spent nine to twelve months negotiating
one
agreement with Motorola—and it was merely a framework for future discussions. “So Andy just tried to do his best to insulate Android from any of that [frustrating bureaucracy]. They didn’t interact with anyone else. They were completely separate.” Schmidt, Brin, and Page even let Rubin build a café inside the Android headquarters on the Google campus that for a while was open only to Android employees.

The idea of a division inside Google that few even knew about was antithetical to its culture. What made Google different from other corporations was that it avoided silos—separate divisions that didn’t interact—at all costs. Schmidt, Brin, and Page had set up the company to actively encourage information sharing. Any engineer could find out what other engineers were working on and even look at the software code with a few clicks of a mouse. Before Google went public—and became subject to SEC rules—Schmidt, Brin, and Page even shared details about Google’s revenues and profits in companywide meetings in front of more than a thousand employees.

Rubin respected Google’s unique approach. But he also understood that if other companies knew what he was working on, they might beat him to the marketplace. “There were plenty of pissed-off Googlers who said we’re not Googley because we’re not sharing,” a former top Android engineer told me. “We had to turn down some very senior people who wanted to see our source code, and Andy had to be the bad guy. So there was a lot of tension.”

Rubin wasn’t just driven by his need to make sure Android moved fast. He knew that producing software for smartphones was vastly different from producing software for the web, which was Google’s primary business. In Google’s web-software world, all products are free and no product is ever truly finished. Juxtaposed against the tyranny of Microsoft and the packaged-software industry in general, this was a truly innovative philosophy. Google would get a product to about 80 percent finished, release it to users, and let their feedback guide the remaining 20 percent of development. Because the software was free, users’ expectations were not as high. And because the software was on the web, the refining could be done almost in real time. There was no longer any need to wait a year until the next release went to stores, which was the way most software was still sold back then.

Rubin knew the cell phone industry viewed Google’s approach to deadlines with horror. When you make and sell physical things such as cell phones, products that aren’t finished in time for the holiday shopping season are catastrophes that waste hundreds of millions in carrier marketing costs, and manufacturer development costs. “I remember some times where Andy would say, ‘We need to get this done by this date,’ and a part of the engineering team would say, ‘We can’t get that done by then,’ and Andy would say, ‘If you can’t get it done, I’ll fire you guys and hire a new team that
can
do it,’” said another former Android engineer.

At most companies such a hierarchical, even militaristic approach to getting things done would be considered conventional. At Google it was so distinctive that it made the Android team feel as if they were revolutionaries. After the iPhone shock wore off and the Android team saw all the things the iPhone didn’t do, its members truly believed that what they were building would be superior in every way, and that they didn’t even need Google to pull it off. “I basically thought there was no way the iPhone could compete with us,” said Bob Lee, a top Android engineer at the time. “I thought Android was going to turn into Windows [because of its vast distribution across many phones] with a ninety-eight percent market share, and that the iPhone would ultimately end up with just two percent market share.”

Rubin encouraged this feeling every chance he got by passing along the perks of his executive-level job to his staff. He was always buying the latest gadgets—cameras, audio equipment, gaming systems, and other electronics—to keep abreast of the latest thinking in his industry. But he rarely kept his purchases long. When he was done, he’d just put them outside his office and send an email to his staff, offering them on a first-come-first-serve basis. Often it was the latest high-end camera or stereo system, worth thousands of dollars. If many on his staff had to be at a conference—say, the Consumer Electronics Show in Las Vegas—he would charter a jet so they could easily get there and back. One year, after Google released the Dream—known by then as the T-Mobile G1—Rubin boosted the Android’s team year-end Google bonus with money out of his own pocket. One engineer said it doubled his year-end bonus.

The downside to all this separateness, however, was that it didn’t endear Android to the Google rank and file any more than Gundotra had with his decision to throw Google’s weight behind the iPhone. As much as the Android team felt as if they could do everything themselves, they couldn’t; and when they needed to work with Googlers on the other side of the wall they’d built, their requests were rarely received warmly. “We’d be like, ‘Hey, we’re doing a phone. Surprise! And we need Gmail on it. Can you help us?’” DeSalvo said. “And they’d be like, ‘Well, we have a two-year software road map and you’re not on it, so, no, we can’t help you.’ So initially we had to use the web API [the connection the public uses] rather than a dedicated API [which would be faster and more reliable]. And it was the same thing for Google Talk, Calendar, and all this other stuff. It was just one nightmare after another, just trying to get basic things done, because no one knew that they needed to support us.”

It wasn’t just the lack of give-and-take between Android and the rest of Google that chilled relations. It was that Rubin’s entire effort at information control wasn’t working well. Every month in 2007 it seemed there was another rumor that Google was building a phone. Googlers were used to being able to keep their products secret because they were typically developed entirely in-house. While Google management shared more than most companies did with their employees, remarkably little of that information leaked. To build Android, however, Rubin needed to work with a myriad of external suppliers and manufacturers. Googlers couldn’t see Android’s code, but some of Android’s external partners could—and some were clearly talking.

Schmidt, Brin, and Page tried to manage Google so that Android seemed to be more integrated into the rest of the company than it really was. But sometimes their actions actually made Android feel even more separate. For example, the trio was so upset about the leaks that at one of Google’s weekly Friday-afternoon gatherings they announced they were launching an investigation to uncover the leakers. For a company that prided itself on a culture of openness and sharing, going after leakers the way U.S. presidents do seemed discordant. Cedric Beust, one of the early Android engineers, said that at some point in the summer of 2007 questions from the staff and the evasive executive responses became so predictable that he and many other Android team members stopped going to the Friday meetings. “It was just too excruciating to have to listen to all this and not be able to say anything,” he said. “The hardest part was hiding my phone [when I was carrying a prototype]. It happened a few times where someone [another Googler] would see me with a phone [prototype] and say, ‘What is this?’ So you would have to have an answer ready. For a while I’d say it was a BlackBerry prototype. Then I’d say it was something we were working on for Nokia. Anything to make it seem as boring and uninteresting as possible.”

For the media, the prospect of a Google phone was a delicious story. It made perfect sense—and seemed downright crazy at the same time. Google had been trying to find a way to disrupt the telecom business almost since its founding. Google’s purchase of unused voice and data transmission lines over the previous three years had carriers convinced Google was planning to become a carrier itself. An entire team at Google was dedicated to finding wireless techniques of routing around the existing telecom infrastructure. Google had been public about its desire to get its applications running on mobile phones. And everyone knew Rubin was working at Google in some capacity. Why would he be there if not to build a phone? Yet Google had just teamed up with Apple to get its applications on the iPhone. Schmidt was on the Apple board. Jobs would be out of his mind with anger if Google was building a competitor to the iPhone.

*   *   *

The secrecy, leaks, and backbiting at Google over its mobile strategy in 2007 meant that when the Android team finally
did
have something to announce, it did not impress. The world was expecting something big. And Google pushed hard for attention—scheduling media conference calls, holding briefings with big software developers, and granting advance access to Rubin and his team to writer John Markoff of
The New York Times
. A profile of Rubin appeared the day before the November 5 announcement, under the headline “I Robot: The Man Behind the Google Phone.”

But Google hadn’t built a phone. It didn’t even have finished phone software. Instead, what Google told the world was that it had built … a consortium of phone makers, carriers, and developers, the Open Handset Alliance. Together, this group was going to team up and turn Rubin’s vision of a better, more unified mobile-phone world into reality. “We are not building a GPhone; we are enabling a thousand people to build a GPhone,” Rubin said.

It was bizarre. It was as if the most interesting and innovative company on the planet had been taken over by United Nations bureaucrats. Schmidt and Rubin touted the consortium’s size—thirty-four companies—and its global reach. They said the software that would be produced from the effort would be free. Phone makers, carriers, and programmers would be able to modify it at will. They said they hoped, but would not require, that manufacturers and carriers would provide a platform for Google so that its own applications such as search and Maps could prosper. The only definitive thing Schmidt and Rubin said about a product was that work was already well under way, that it was called Android (no one had known what the effort was being called until then), and that phone maker HTC would release a phone with that software on it in a year’s time.