New York Times
March 2005

In the post-dot-com boom era, the mind-set among many, if not most, Silicon Valley entrepreneurs is to position a start-up company for a quick and lucrative sale to an industry giant rather than wait for the uncertain payout of an initial public offering.

Keerti Melkote, however, rejected that strategy -- and has grown accustomed to people asking if he has lost his mind.

Mr. Melkote, a founder of a three-year-old start-up called Aruba Networks, which sells wireless networks to corporate customers, says he recognizes that he and his partners might have turned away a once-in-a-lifetime chance at instant riches when executives from Cisco Systems, the largest maker of Internet networking equipment, came knocking last November. Cisco wanted to talk about acquiring Aruba, which has its offices here.

''We basically told them, 'There's not a number you can come up with that will be attractive enough to convince us to sell,''' said Don LeBeau, Aruba's chief executive, who had served as a senior vice president at Cisco in the 1990's.

In January, Cisco announced that it had paid $450 million to purchase Airespace Inc., Aruba's main rival in the market to build and manage wireless networks inside corporate offices.

The $450 million ''is a decent payday -- it's nothing to sneeze at,'' said Mr. Melkote, who oversees product development and marketing.

''But,'' he added, ''that's where you need to have the conviction that your product is better, the market is there, our team has the ability to execute.''

That kind of confidence is much less evident in Silicon Valley now than it was five years ago.

The bursting of the dot-com bubble humbled all but the most big-headed entrepreneurs. And the stock markets, while hungry for high-growth companies, now demand that most start-ups show some profits or other measures of success before going public.

''For the average company in Silicon Valley, the strategy is to be bought rather than go public,'' said Paul F. Deninger, an investment banker and chairman of Broadview International, who advises start-ups on both initial public offerings and acquisitions.

The passage in 2002 of the Sarbanes-Oxley Act, the corporate reform law, has also trimmed back expectations, said Michael J. Sullivan, a lawyer in San Francisco who advises start-up companies. Under that law, chief executives and chief financial officers are personally responsible for their companies' financial statements.

''In the past, the dream of every young company was to go public,'' Mr. Sullivan said. ''Today, with Sarbanes-Oxley, going public means as many problems for a young company as it does glory. The trend is that fewer companies are going public and more are choosing acquisition as an exit strategy.''

Still, for some brave -- or perhaps foolish -- souls, the gamble for the initial public offering is irresistible. ''For the extraordinary company, the I.P.O. is still a very real possibility and the rewards are huge,'' Mr. Deninger said.

Aruba's executives and venture capital investors are clearly eager to take that risk, Mr. LeBeau said.

They were unanimous in deciding to politely brush off Cisco's interest -- though the company's 170 employees, many of whom have not enjoyed the past financial rewards of those running the company, were less certain.

''We believe that a truly big payout will come in the form of an I.P.O., or an acquisition once we've gone public and established a marketplace,'' Mr. LeBeau said.

Aruba's connection with Cisco goes beyond an acquisition interest. In addition to Mr. LeBeau, the three founders -- Mr. Melkote, Merwyn Andrade and Pankaj Manglik, who has already left the company to found another start-up -- were all former Cisco employees. Mr. Melkote and Mr. Andrade worked together there to design switching equipment. Mr. Melkote left in 1997 to work with friends at another start-up and Mr. Andrade left in 2002 to help found Aruba.

Aruba is one of several companies, including Airespace and Symbol Technologies, that are developing less expensive, more secure wireless Internet networks, called local area networks or LAN's, that allow businesses to transform an office building into one giant wireless hot spot, all managed by a central computer. Aruba's greatest advantage over its competitors, some industry analysts say, is that it offers the most secure wireless systems.

''The centralized wireless LAN represents a great architectural leap forward,'' said Craig Mathias, president of the Farpoint Group, an industry consulting firm in Ashland, Mass. It also represents a market that Cisco has been eyeing for at least two years.

Cisco, based in San Jose, Calif., first approached Aruba in early 2003, when it had just started selling its wireless systems. Leading the Cisco team was Dan Scheinman, the executive charged with keeping tabs on new rivals that might present a competitive challenge or an acquisition opportunity.

''We wanted to see what was there, and whether or not there was a real market and something interesting for us,'' Mr. Scheinman said. As he tells it, the two sides talked, but did not go further because he was not persuaded there was a large enough market for Cisco to exploit. ''Obviously I was wrong about that,'' he said.

With the Airespace acquisition, Cisco intends to extend its domination over the networking market to include the world's leading wireless LAN offering. According to Mr. Melkote, the meeting may have given the Cisco team precisely what it wanted.

''We were a little naïve,'' said Mr. Melkote, who was 33 at the time. ''We gave them our product to examine so they could see it was real.'' He shook his head and laughed at his own foolhardiness. ''Cisco came out with a centralized wireless product around a year later that looked remarkably like our product,'' he said.

Mr. Scheinman objected to that suggestion as an unfair ''attack on our integrity.''

David Callisch, Aruba's communications director, said that the episode spoke more to the company's inexperience than to the actions of an aggressive foe. ''Our founding team was very young and very enamored with all the attention they were getting from the likes of Cisco,'' he said. Later that year, Aruba hired Mr. LeBeau to run the company.

That Cisco tried to initiate talks again late last year was surprising, said Mr. LeBeau. It posed the question, he said, of why the company felt compelled to ''spend $450 million to acquire what they've been trying to build.''

Mr. LeBeau is convinced that Airespace was Cisco's second choice, and he even says that one of his former Cisco buddies told him so. But Mr. Scheinman said Cisco had been in talks with Airespace since last August, before the company reached out to Aruba a second time.

For Aruba's top executives, the disadvantages of life as a small unit of Cisco -- constantly battling for resources and attention -- outweighed the advantages of Cisco's brand and marketing muscle.

''I saw that it would have been impossible for us to fulfill our dream inside of Cisco,'' said Mr. Andrade, Aruba's chief technology officer. ''I feel we're really on the cusp of this taking off, but know from my experiences inside Cisco how difficult it would have been to innovate.''

Aruba's goal is not to beat Cisco so much as to carve out a sizable share of the corporate market for wireless networks, a market that Mr. Melkote estimates will be worth more than $10 billion in the next five years.

''Even if it's only a $5 billion market, and even if we only get a 20 to 40 percent share of that, we can go public on that,'' Mr. Melkote said.

Photos: Merwyn Andrade, Keerti Melkote and Don LeBeau of Aruba Networks have all worked at the rival company that tried to buy theirs, Cisco.; Kevin Huey, a network engineer, runs a test of the type of wireless systems Aruba sells to businesses.