Grand Junction Networks

How can you call the news good - even if it means you're going to be rich - when the bearer seems to be in such pain? How can you call the news bad - even if it hurt - when it makes so much sense?

In September 1995, Grand Junction Networks was growing out of its skin. The market for its products - digital switches that made Ethernet networks run 10 times faster than anyone had thought possible - was exploding. Projected annual revenues were nearly $32 million, more than four times as much as the year before. The company had moved from a shabby warehouse in the drive-by town of Union City to a respectable building in Fremont.

As the company's senior parts buyer and planner, Debra Pelsma faced demands on her time that were as insatiable as the demand for Grand Junction's products. Fourteen-hour days were common. "My husband had been through a startup," she says. "He understood that I was getting wrapped up in it. We were going to kick butt. It was the price of a dream, a part of the equation."

At the center of the equation - weaving the dream - was Howard Charney. Charney, then 47, was already a notable figure in Silicon Valley. He was a cofounder of 3Com, a leader in computer-networking systems. But after 10 years there, after having fun at almost every big job in the company, he had started having less fun - and walked away.

The move was consistent with his character. Charney is as soulful as he is successful. His dark eyebrows rise dramatically above his glasses, like the dorsal fins of a sailfish. His voice is full yet gentle, like the rocking of a boat.

Pelsma, then 42, met Charney in the summer of 1993 and immediately became a disciple. "He was so optimistic, so sincere, so genuine," she says. "I decided I'd follow him anywhere." Charney is "the kind of guy people walk through walls for," confirms Kathryn Gould, one of Silicon Valley's top-producing female venture capitalists and the lead investor in Grand Junction.

Charney left 3Com for complex reasons. As the company got bigger, the turf battles got more ferocious, and he just didn't want to fight. "A corporate gladiator I am not," Charney says. "I walked into [CEO] Bill Krause's office and said, 'I'm going to retire.' He said, 'Should I try to keep you?' I said, 'No.' He didn't say much else. We'd been together a long time."

Charney shed 3Com - but he couldn't shake the entrepreneurial bug. "I began to miss the really smart people, the great human beings," he says. "I was no longer tuned in."

So a year and a half later, he started calling these smart people. Would they like to drop by and talk? In his living room and kitchen, high-octane thinkers, six or seven at a time, began meeting every few weeks. They threw out all kinds of wild ideas, most of which they dismissed. "Then," Charney remembers, "we're sitting around one night - this is in September 1991 - and one of the guys says, 'Maybe we can make Ethernet go 10 times faster.' We spent the next several hours trying to knock the idea down. It seemed so simple."

The idea was simple - and brilliant. Charney explained it to Gould, then a partner at Merrill, Pickard, Anderson & Eyre, a premier West Coast venture-capital firm. Gould got very interested. Charney got more than $1.5 million in seed money.

Grand Junction was born shortly thereafter - in February 1992. The company exhibited few pretenses of prosperity. The office in Union City was shabby, even by warehouse standards. Margot Gangola, now 34, took one look and said no the first time Charney invited her to join. "Howard had a great reputation," she says. "But the company seemed too small, too high-risk." Charney persisted. He wanted her to be on his engineering team. Eventually Gangola relented. "It was incredibly hard work," she says. "I knew I'd live through long days and nights for at least 18 months, just to get up and running. But we had an amazing team."

More venture capital came in - and so did revenues. First-year sales were $4.5 million. In the second year, 1994, revenues increased to $7 million. Things got even hotter in 1995. Charney was beginning to tell people, "We think this thing is going to take off."

Other people were thinking the same thing. Back in 1993, when Grand Junction was shipping its first product, Cisco Systems was already a Silicon Valley powerhouse. And every year since, Cisco had grown richer. By 1995, the San Jose-based giant had revenues of more than $2.2 billion. But Cisco was missing something - low-end, fast-Ethernet switching products. That's what Grand Junction built.

Quietly, secretly, Cisco started looking at Grand Junction. How much business would the young startup do in 1996, Cisco's financial wizards asked themselves, if it operated as part of Cisco? Their stunning answer: $119 million. That was 17 times what Grand Junction had made in 1994.