Shy, Unassuming, and Controversial

"I want people to take the risks necessary to drive innovation," says Sun's CEO.

By Al Riske

9.Dec.04--Scott McNealy is feeling good. The company he cofounded has just posted a second quarter of growth. Two in a row. Funny, if you listened to the conventional wisdom, that was never going to happen. But then McNealy has never been one for conventional wisdom.

"If your strategy isn't controversial," he says, "you have zero chance of making money."

In 1982, Sun bet everything on open interfaces and open standards. It was, back then, a completely contrarian and counterintuitive business model.

"People couldn't understand how we expected to survive if we couldn't lock customers in," McNealy recalls. "I couldn't understand why customers would allow themselves to be locked in."

You have to understand that McNealy grew up near Detroit, the center of the automobile industry in the U.S., and his dad was an executive at American Motors. Standards were the norm -- "gas pedal on the right, brake pedal on the left" -- and yet car makers were still able to differentiate themselves.

Customers benefited, McNealy says, because they didn't have to take driver's education all over again if they wanted to trade in a Ford for a Chevrolet.

In 2004, it's no longer considered radical to say that computers should work together no matter who made them. So, McNealy and company are finding other ways to stir things up.

"In the tech space, the conventional wisdom has never been more conventional than it is right now," he says.

McNealy sees companies all over the world doing the same thing -- pouring money and manpower into what he calls "custom jalopy" data centers -- and it hardly seems wise to him.

"I've never walked into a data center and said, 'Hmm, this looks just like the last one.' No two are alike," he says. "That's the wrong answer."

The right answer? "Standardized, consistent, and integrated -- that's the way to go," McNealy says. "Map your company's workloads to industry standards such as Java Web services instead of investing in customized one-off solutions.

"We're turning heads with our N1 Grid program -- compute power for one dollar per processor per hour," he adds.

Sun has set up two 1000-CPU grids, signed a number of beta customers, and captured the attention of many others. Biotech companies doing drug research, Wall Street firms running market simulations, investment banks running risk-analysis programs, and startups that can't afford their own supercomputers are all potential buyers of extra compute cycles.

The trick will be convincing customers to move to the pay-as-you-go model, especially customers accustomed to owning their own systems and running them within their own four walls. In short, the issue is not technical but cultural.

McNealy understands that, but he simply smiles and says, "We just have to get people comfortable with the fact that their money is safer in the bank than in their mattresses."

His real aim, though, is for Sun to sell compute power wholesale rather than retail. "And we're already doing it through companies like ACS, EDS, and Salesforce.com," he says.

So Sun has been reinventing itself. Again. And, just as before, there are skeptics. McNealy seems to like it that way.

An amateur ice-hockey player, he has always had a strong competitive streak. He knows, too, that you don't skate to the puck but to where it's going.

So he has steered Sun from the desktop to the server, from the server to the Web.

"We've still got to finish that transition," he says. "Then we'll have the next big shift, which will be to a computing environment with components that network to each other as opposed to big monolithic server infrastructure environments."

McNealy gets a lot of free advice and public performance reviews from analysts and business writers when Sun moves into new areas that threaten to cannibalize sales from other parts of the business -- offering small, low-cost servers that might take sales from high-end machines, for example.

At the same time, others say he was too slow to do so.

He shrugs it off. "Look at it this way," he says. "If you were a ballplayer, wouldn't you rather be inducted into the Hall of Fame by the voting of your peers -- the players who really know the game?"

Ask McNealy about the mistakes he's made, though, and he'll tell you: Hiring too many people and signing too many leases during the boom years of the 1990s. Underestimating the longevity of 32-bit microprocessors for use in Web servers. Paying too much for certain acquisitions. Not offering Solaris bundled with x86 hardware seven years sooner.

"I have yet to meet the perfect human," he says, "and I'm certainly not it."

McNealy says his job is to take the blame for missteps and spread the credit for successes.

"When I admit mistakes, it lets my team know that it's okay to take risks," he says. "I want people to take the risks necessary to drive innovation."

He makes it clear, though, that Sun's successes have always far outstripped the setbacks.

"You have to have a wildly different strategy and you have to be right," he says. "It's that second part that gets tricky."

Sun was right, he says, when it moved to a 64-bit chip architecture eight years ago, well ahead of the pack. That put the company in a position to capitalize on the Internet boom and put more than $7 billion in the bank.

More recently, the company made another bold move in the processor space. Several years ago, Sun started experimenting with a multicore, multithread design that actually runs slower (and cooler) but still gets more work done.

"It was absolutely the right decision," McNealy says. "Now Intel is having to say, 'Whoops! We goofed,' and rethink its entire product line."

Sun was right, too, when it decided to "double down" on the Solaris operating system, McNealy says, even as Linux was being hailed as the wave of the future.

"The new Solaris 10 release is far and away the best operating system in the history of computing," he says. "And customers who have been trying it out agree."

These and other milestones have culminated in a strong end-to-end product line, and Jonathan Schwartz, Sun's president and chief operating officer, credits McNealy's courage and commitment.

"At a point where everyone else was chopping R&D left and right, Scott took a pretty controversial stance that basically said, 'Customers care about R&D.' And I think the single-biggest reason -- clear and away beyond anything else -- that customers love talking to us right now is because Scott preserved the R&D roadmap."

Over the past couple of years, McNealy could be called a contrarian simply because of his relentless optimism. He certainly recognizes the many perils of the competitive landscape, but he chooses to emphasize the positive.

"There's just something about Sun's mission, something about the cause -- the belief that we can bridge the digital divide, that we can provide choice, that we can open up the computing industry, that we can simplify the world, that we can bring people Web services in ways they never thought possible," he says.

McNealy points out that Sun systems are used for modeling weather systems, mapping genetics information, running the world's financial systems, treating patients, and teaching students in far-flung locations.

"Just look around at how people use our equipment, and there's something that can get just about anyone excited," he says.

People who have worked with him for years will tell you that Scott McNealy is not an easy man to get to know. He can be fierce when he needs to be, but he's also fun-loving and genuinely charming.

Anyone who has seen him on stage knows there's a touch of the entertainer in him. At the end of the day, though, he'd rather go home and have dinner with his family than attend a cocktail party.

A little-known fact: McNealy was described in his high school yearbook as shy and unassuming.

"And I still am," he says.

He's never shy about Sun, though. If anything, he's prouder than ever of the company he cofounded nearly 23 years ago. And why not? With back-to-back quarters of revenue growth, he has it moving in the right direction again.

"There's lots more to do," he says, "but this feels really good for a change."