"Is this interesting to Google?" That's what Andy Rubin was asking Larry Page. It was a spring day in 2005, and the two were in a conference room just off the main lobby at Google's headquarters. A simple yes and Rubin would have walked away happy.
They had met three years before, when Rubin was about to launch a smartphone he'd invented called the Sidekick. At the time, Google was just an up-and-comer, trailing AOL and even Lycos in traffic. But Rubin, a well-known Silicon Valley player, chose Google as the Sidekick's default search engine. Page was flattered by the unexpected endorsement. So when Rubin called out of the blue and requested this meeting, well, Page couldn't say no.
The Google cofounder arrived late, as usual. Rubin walked to the whiteboard and began his pitch. There were nearly 700 million cell phones sold each year compared with fewer than 200 million PCs — and the gap was widening. Increasingly, he said, phones were the way people wanted to connect with each other and with everything else. Yet the mobile industry was stuck in the dark ages. Unlike the Web, where open standards had fostered a multitude of cool companies and applications, mobile was a tyrannical, closed system, repelling all innovators and disrupters who tried to gain entrance.
Rubin said his startup, called Android, had the solution: a free, open source mobile platform that any coder could write for and any handset maker could install. He would make his money by selling support for the system — security services, say, or email management. Android would have the spirit of Linux and the reach of Windows. It would be a global, open operating system for the wireless future.
Rubin didn't want money from Page. He already had funding. What he wanted was Google's imprimatur — even an email from Page would do. Rubin figured he could attract more VC funds with the search giant on board, possibly with a hint that Google might be interested in developing its own branded phone. He pulled out a prototype.
Page picked up the device. He had been personally frustrated yet fascinated by the mobile market for years. He already knew the numbers — he didn't need Rubin to tell him how many PCs and mobile phones were out there. He also knew that it added up to a massive problem for Google.
The desktop metaphor was fading. Phones were going to replace PCs as the main gateway to the Internet, and they were going to do it soon. Why would consumers tether themselves to a PC when phones were growing more and more powerful — and were cheaper, too?
But because cell phones ran on different software, had less memory, and operated under the constraints of pay-per-byte wireless networks, the mobile Web was a stripped-down, mimeographed version of the real thing. Reading and surfing and — more to the point — viewing Google ads was a slow, stultifying chore. Even worse, a second-class Web could derail Google's grand strategy. The company was trying to worm its way deeper into users' lives by hosting applications and personal files on Google servers, then dishing them out to the always-connected consumer whenever and wherever needed. That was easy on PCs, but phones didn't play nice with the cloud. Google dominated the Web today, but tomorrow might be a different story.
Working the problem had been a nightmare. Google engineers had a closet overflowing with mobile phones to test the company's wireless applications — mobile Google, Blogger, search over SMS. There were dozens of operating systems to navigate, a mobile Tower of Babel completely at odds with the easy access and universal language of the Web.
What worried Page most was that the only firm from the PC world that seemed to be successfully navigating the mobile labyrinth was Microsoft, one of Google's biggest rivals. The Windows Mobile platform had less than 10 percent of the US smartphone market, but it was growing fast. Microsoft's system, however, was the ugly stepsister of what Rubin was proposing: Redmond executives cared less about opening up the Net to mobile users than about tying the mobile operating system into its desktop dominance. A decade ago, Microsoft had underestimated the growth of the Web and then lost control of it to Google. Now it looked like it was Google's turn to be caught flat-footed.
If Google had it bad, users had it worse. Every year since 2002, the wireless sector managed to place at or near the top of the Better Business Bureau's tally of the most complained-about industries. Americans would rather do business with a used-car salesman or a collection agent than with a customer service rep for, say, T-Mobile or Motorola. And who could blame them? The plans were expensive, pricing was complex and capricious, and the phones never lived up to expectations. Constant innovation, the first principle of Page and Rubin's world, was anathema to phone companies. There had to be pent-up demand out there for something better.
So was Rubin's pitch interesting to Page? Absolutely. But he didn't want to stick his logo on Rubin's phone. Or write a supportive email. He had a better idea: Google would buy Android.
Rubin was floored. He had come in looking for an encouraging word and left with the biggest payday in his life. (The eventual purchase price was estimated at as much as $50 million.) Now all he had to do was live up to his own hype.
"Google's model is to build a killer app, then monetize it later," Rubin says. We're sitting in another conference room across the street from where he and Page struck their deal three years ago. The building, which houses Google's mobile division, is Rubin's domain now. There's a self-piloting model helicopter bearing an Android logo parked in the hall — Rubin builds them in his spare time. Beyond are floors of people who think of nothing but the cellular future of their employer. In the lobby, a flatscreen TV shows a spinning globe with animated flares erupting wherever people are using Google to search from their mobile phones. This fall, when the first Android phones hit the market, those flares will presumably flame even higher.
Rubin is tall and skinny and a casual dresser even by Google standards. He's 45 but seems younger. Sitting with one leg tucked beneath him, he explains the mission of the Googlefied Android to me, but I barely follow the words. I'm staring at his phone. It's clearly a demo — black, scuffed, covered with fingerprints; most of the face is taken up by the screen. Rubin absentmindedly slides it around the big wooden table, then picks it up and shifts it from hand to hand. It's maddening. All I want to do is get a closer look at his killer app.
After Google bought Android in July 2005, Silicon Valley pulsed with gossip and speculation about what the search giant was planning. Everyone figured Apple had a phone in the works and assumed Google must be developing one too. Rubin and his cofounders, Rich Miner, Nick Sears, and Chris White, weren't talking. "Trying to guess Google's next move recently replaced digging through Steve Jobs' garbage ... at the top of our weekend activities list," wrote tech blog Engadget. When Apple unveiled the iPhone last summer, expectations for a gPhone — could it be called anything else? — grew even more feverish.
But when Google finally broke its silence in early November, there was nothing about a gPhone. Instead, there was a press release. Thirty-four companies — firms like Texas Instruments, Intel, T-Mobile, and Sprint Nextel — were joining Google to build a wireless interface based on open source Linux software. The group dubbed itself the Open Handset Alliance. Competitors sighed in relief. This was how Google planned to shake up the nearly trillion-dollar global wireless market? A consortium?
"Their efforts are just some words on paper," remarked Steve Ballmer, CEO of Microsoft, at a conference in Japan. "Another Linux platform," shrugged the CEO of Symbian, the dominant smartphone operating system outside the US.
A week later, Google upped the ante. The company put a free Android software developer's kit on its Web site and announced the Developer Challenge, with $10 million in prize money to be parceled out to the creators of the best applications for the new system — a great social networking tool, say, or a handwriting recognition program. The Challenge was an open call; anyone was invited to take a shot.
Those hoping for a new gadget to rival the iPhone finally understood that Google had something radically different in mind. Apple's device was an end in itself — a self-contained, jewel-like masterpiece locked in a sleek protective shell. Android was a means, a seed intended to grow an entire new wireless family tree. Google was never in the hardware business. There would be no gPhone — instead, there would be hundreds of gPhones.
HTC, Motorola, and LG all announced plans to market new Android phones in a multitude of shapes and sizes, each with different software options. Android was a fully customizable system. Any application could be removed or swapped out for another. Even the few programs that Google was creating from scratch — an email app, a contacts manager — could be replaced with third-party software that did the same thing. Google didn't care how any individual model was pimped out as long as the hidden Android DNA was there underneath, keeping everything tied to the Internet and running smoothly.
As soon as programmers started playing with the emulator, they saw how big Google's ambitions were. The company was trying to make programming for a cell phone analogous to programming for a PC or the Web. Coders were told that their applications would have constant access to the Net, not the usual mobile hurry-up-and-wait feel. Working with the cloud — enabling programs to push or pull info to or from the Web — was a must. All Android phones would know where they were at all times, either by tapping into onboard GPS or by cross-referencing cell towers using a proprietary database owned by Google. And applications would be allowed to share information, which at the simplest level meant the kind of copy-and-paste functionality across all programs that cell phones currently lack.
Even better, at least in a developer's eyes, the Android team had violated an essential tenet of the wireless industry: that users are too dumb and dangerous to be trusted with downloadable software. Engineers who write for just about any mobile operating system today have to spend time and cash obtaining security keys and code-signing certificates. Android would allow any application to be installed and run, no questions asked.
