Hard Cell

If space inside the phone was precious, weight was even dearer. The original pdQ Smartphone had weighed 285 grams -- more than four PowerBar snacks -- and that heft made it laughably unattractive. The new phone would weigh nearly 25% less, vowed Koerper and senior engineer George Wiley. This time, the battery couldn't weigh more than 43 grams. The flip element on the phone was allotted just 16 grams. And when de Lara found a new microphone model that weighed just 0.3 grams instead of 0.4 grams, everyone cheered.

At times, even the basic laws of physics seemed to conspire against Koerper's team. Mobile phones work by detecting extremely faint signals in the air and then amplifying and filtering them to the point that they become recognizable conversations. That's nice, but a PDA doing even the simplest task emits electrical signals that are 1,000 times stronger than the incoming calls a phone module is trying to hear. Without an artful interplay of shields, amplifiers, and filters, users trying to make a phone call will hear nothing except the chaotic hum of the PDA in action.

One of the most challenging aspects of the final design called for the phone's circuitry to use filters no larger than a capital O on this page. Suppliers were accustomed to making those parts several times that size. "When I first told them what we wanted, they just laughed and said, 'No way,' " engineer Marimuthu Sivakumar recalls. "But we eventually found someone who could do it."

In hindsight, much of the early engineering work was planned imperfectly -- if it was planned at all. Team members meddled in one another's areas, changed their minds a half-dozen times about things as fundamental as battery design, and conducted some of their most important meetings during 90-second hallway chats.

Yet team members said that such improvisation is exactly the right way to guide a project in a fast-moving industry. "No matter how much you think you know at the start, you're going to face surprises almost every day," Koerper explains. "Some things that seemed impossible are going to become practical. And some things that seemed easy are going to be problems. It wasn't until nearly a year into the project that I really knew all of the pieces were going to fit together inside the phone."

By autumn 1999, Koerper felt confident enough in the phone's progress to start thinking about its appearance. He and Halkola wanted colors that projected such messages as "high-end," "serious," and "cutting edge." The body of the phone should be charcoal gray or black. The accent elements could then be a metallic silver. To test their ideas, the two men traveled to Selkirk, New York, where GE has a huge plastics lab.

"They've got millions of colors there," Koerper says. "They can mix up anything you want. It costs thousands of dollars a color, but it's mesmerizing." As he played around with colors, he decided to scrap the bright silver for a more subdued gunmetal gray. At another point, Koerper says he told GE technicians that they were close, but that he wanted something more, um, masculine. "I figured that they would look at me as if I were nuts," he says. "But they didn't. They came back a few minutes later with exactly what we wanted."

Business Model: Underpromise and Overdeliver

All the while, Koerper fine-tuned the Smartphone's business projections. In the mobile-phone business -- as in most high-tech sectors -- no one really knows how well a new product will sell until it hits the market. But executives must make their best guesses anyway. Koerper's spreadsheets in late 1999 envisioned selling more than 250,000 units of the phone over an 18-month life span, pricing them at about $560 apiece. That would be small fare in an industry that sells 80 million phones a year in the United States alone. But Koerper thought that the Smartphone could be nicely profitable even at that level.

Besides, Koerper would later explain, "The original pdQ phone was just dying in the marketplace back then. There was no sense in raising top management's suspicions by claiming we were going to do great things with this model. I felt that we would be much better off if we underpromised and overdelivered."

For senior executives at Qualcomm, though, the picture was much muddier. Two simpler and cheaper mobile-phone prototypes were racing to market at the same time as the Smartphone. Typically, Qualcomm launched only one model at a time; it had never done three at once. Multiple launches could splinter the company's sales-and-marketing focus, executives worried. In an effort to pitch too many different phones at once, the company might fail to deliver a strong message for any of them.