Sun: A CEO's Last Stand

Scott McNealy knows he made many mistakes. Is it too late to recover?

Talk with ever-voluble Sun Microsystems Inc. (SUNW ) CEO Scott McNealy, and you may hear one of his favorite quips: "Conventional wisdom doesn't contain a whole lot of wisdom." He believes it because of his own experience. Consider 1995: All of Sun's competitors -- Hewlett-Packard (HPQ ), IBM (IBM ), and Digital Equipment Corp. -- were busy developing new servers to run the next version of Microsoft Corp.'s Windows software. Wall Street pundits begged McNealy to show some common sense and do the same. But he refused, instead cranking up his investment in Sun's own software, called Solaris. What happened next made McNealy look brilliant. Rivals couldn't match the speed, reliability, and security of Sun's servers. As the tech boom took off, Sun's boxes became the must-have gear for thousands of Internet startups and financial firms. Sales soared; profits exploded.

Six years later, as the boom of the late 1990s came to a crashing end, Wall Street had more advice for McNealy: Batten down the hatches for the storm ahead; slash research; lay off staffers; and get serious about low-cost products. Once again, McNealy held his ground. But this time, he was dreadfully wrong. Sun's sales have tumbled 48% in the past three years, it has lost a third of its market share -- and it continues to head south even as its rivals ride the economic recovery. Its stock, which reached $64 in 2000, trades at about $4. No other major player has been weakened as much during the tech downturn. "Right now it looks pretty grim," says Geoffrey A. Moore, author of several tech-industry books, including Crossing the Chasm.

It's a classic management tragedy, and to a striking degree the responsibility lies with the 49-year-old McNealy. His greatest strengths -- the uncompromising determination, sharp-tongued irreverence, and unblushing idealism -- turned out to be critical flaws. Through interviews with 38 current and former Sun executives, including nine departees on the record, BusinessWeek has learned that as Sun's situation deteriorated, McNealy was bucking not just the counsel of outsiders but also that of his own lieutenants. After the tech industry went into its long slide in late 2000, virtually his entire management team, including Chief Scientist Bill Joy and President Edward J. Zander, pleaded with McNealy to scale back his vision and adjust to meaner times.

Time and again, McNealy refused. An economics major during his days at Harvard University, he was convinced that the economy would snap back quickly from its slump, insiders say. Plus, he believed that the Net was so critical to companies that they couldn't hold off buying gear for long. "The Internet is still wildly underhyped, underutilized, and underimplemented," he said in early 2001. "I think we're looking at the largest equipment business in the history of anything. The growth opportunities are stunning." Preparing for the next upturn, he felt, was much more important than whittling expenses for a brief lull.


As the tech wreck went from bad to worse, McNealy's contrarian instincts kicked in. After all, he had been right to ignore the consensus within Sun's ranks before. In the 1980s, he overruled execs who were skittish about dropping Motorola Inc.'s microprocessors for chips developed by Sun -- a move that paid off in a big way. This time, as his team urged him to cut back, he felt the stakes were even higher. He was determined to fight off what he thought were short-term thinkers, particularly on Wall Street, so that Sun could be preserved as an innovative force. Although he had thought about quitting during the boom, McNealy recommitted himself to Sun in late 2001, convinced that his credibility, experience, and sheer nerve were what the company needed during its darkest days. "I'm here, and I'm not going away. This is a really tough situation, and we're going to get through this," he told staffers at the time, according to former Executive Vice-President Larry Hambly.

It would be nice to think that McNealy was doing the right thing. After all, here was a chief executive who was taking the heat to protect his employees, fund R&D, and plan for the long term. Alas, it was not to be. He badly underestimated the severity of the downturn and dismissed customers' desire for low-end servers. As time wore on, the losses piled up, and McNealy's high-minded resolve began to look to others like simple-minded obstinacy. One by one, his team lost faith and departed. All told, almost a dozen of McNealy's most trusted lieutenants have left over the past three years, including Zander, Joy, and John Shoemaker, chief of the server business. Like many others, Masood Jabbar, Sun's longtime sales chief who retired in 2002, says he admires McNealy's courage. But the standoff became counterproductive. "The fight just didn't seem worth it anymore," says Jabbar. "It was an untenable situation."

Now some investors believe it's time for McNealy to follow his former execs out the door, or at least give up the CEO post and retain only a chairman's role. Says analyst Andrew Neff of Bear, Stearns & Co. (BSC ): "It's pretty standard that if the ship keeps going toward the iceberg, you change the captain."

But this captain likely will remain at the helm for the foreseeable future. Two Sun directors who agreed to interviews for this article, M. Kenneth Oshman and Naomi O. Seligman, say the board is squarely behind McNealy. "There's no plan for Scott to step down. I think we've got a great leader," says Oshman, CEO of networking player Echelon Corp. (ELON ).

As for McNealy, he says he still has what it takes to bring Sun back. "Maybe it's time to get rid of me," he says. "But this company has a lot invested in training and developing me. I have 20 years' experience. I'm 49 years old. I'm in good shape. Healthy. Lot of energy. Lot of wisdom. Relationships around the world." He seems remarkably unperturbed by the withering criticism of the past few years. Although he admits to some mistakes, he's just as acerbic and cocky as ever. He's not prone to self-doubt, or even much self-reflection. "I don't do feelings," he has said. "I'll leave that to Barry Manilow."

Instead, McNealy is focused on turning Sun around with what he calls "disruptive innovation," the same approach that has saved it so many times before. While most rivals make plain-vanilla computers and slug it out on price, Sun's plan is to change the rules of the game. At the high end of the server market, Sun is developing "throughput computing" chips that can handle dozens of tasks at the same time. At the low end, Sun servers built around Advanced Micro Devices Inc.'s (AMD ) inexpensive chips will handle not only processing tasks but also the basic networking that rivals' boxes can't. And its pricing approach is something no server company has dared try before: It's planning to give away low-end servers to customers that agree to buy its software for several years. "We have a maverick strategy," says McNealy. "I think there's a huge opportunity right now."

He insists that concerns about Sun's future are vastly overblown. He points out that unit sales of the company's servers have risen more than 20% in each of the past three quarters. Despite huge write-offs, Sun has $7.5 billion in cash and investments, including nearly $2 billion from a landmark settlement with Microsoft in April. And he pledges that he's getting serious about whacking away at Sun's bloated costs, having laid off 28% of his workforce in three rounds.

While McNealy's plan may look good on paper, it will be difficult to pull off. The new strategy calls for Sun to move in two directions at once: build bare-bones servers while also inventing cutting-edge technologies. Those are diametrically opposed pursuits, like trying to be Wal-Mart (WMT ) and Gucci (GUC ) at the same time. McNealy contends that Sun is more focused than major rivals. Dell Inc. (DELL ), for instance, sells printers and digital music players, while IBM gets half its revenues from services. "We're not doing digital cameras. We're not doing printers," says McNealy. "We're fundamentally focused, much more so than any company I see out there."


But Sun's competitors are focused in other, potent, ways. Dell concentrates on efficiency and low costs, spending a mere 2% of sales on R&D. That'll cause trouble in basic hardware for Sun, which invests 17% in R&D. At the same time, Sun will struggle to out-innovate larger, and deeper-pocketed, rivals such as IBM. And Sun's track record of diversification is lousy. The reality is that every major initiative to move beyond high-end servers over the past decade has failed. "Scott's a smart guy, but I don't see a way out for Sun," says Kevin B. Rollins, chief executive of Dell. "Will they disappear? No. But most of the customers we talk to are looking for reasons to use less of their products."


The bottom line is that McNealy's moves probably come too late to resurrect the icon-turned-also-ran. When the company reports earnings for the fourth fiscal quarter on July 20, analysts expect slight improvements, including stronger low-end sales and narrower losses, excluding special charges. But revenues are forecast to be about 5% below the year-earlier quarter. And for the fiscal year ended June 30, analysts expect revenues to have slipped 4%, to $11 billion, and net losses to have hit $1.2 billion, excluding any gain from the Microsoft settlement. Most industry experts expect larger and stronger competitors, including ibm, Dell, and Hewlett-Packard, to continue making off with blue-chip customers, relegating Sun -- and McNealy -- to side stage. "Scott is kind of like Moses," says Paul Saffo, director of the Institute for the Future, a Menlo Park (Calif.) think tank. "[He] led the world to the land of milk and honey, but he got left behind."

McNealy's willingness to buck popular opinion dates back to his days growing up outside Detroit. His father was vice-chairman of American Motors Corp., and Scott was an accomplished kid -- honors student, standout clarinetist, captain of the tennis team. One summer, he worked in an auto-parts factory. When the United Auto Workers at the plant went on strike, McNealy didn't think twice about crossing the picket lines -- despite bomb threats and jeers from angry union members. "It seemed incredibly stupid," he said. "I couldn't see how highly paid uaw workers were helping their cause" by losing the company money.

He didn't rush to Wall Street or a tech giant after getting his MBA from Stanford, either. Instead, after "majoring in beer and golf," as he describes it, he helped found Sun in 1982, taking on the workaday manufacturing duties, while Joy and others cooked up the sizzling technology. After being named CEO in 1984, McNealy quickly showed a bent for making brash bets to stave off disaster. Boosting investments in Sun's Solaris software in 1995 was the boldest. Then, as the Net boom began, McNealy helped turn Java into Sun's own tsunami. Because Java makes it possible to write Net programs that run on any kind of hardware, developers jumped on the bandwagon, seeding the market for Sun's gear.

The result was a frenzy of sales. By 2000, Sun was turning in scorching growth of 50% a quarter -- faster than Microsoft (MSFT ), Intel (INTC ), or Dell. It racked up gaudy profits, nearly $2 billion in 2000. For that short-lived era when quality was an imperative and price was no concern, Sun ruled.

But Sun's success laid the seeds for the trouble that would follow. The company poured billions into R&D for a mind-numbing galaxy of projects -- from UltraSPARC chips for closet-size servers to software for smart credit cards. Sun hired 7,000 new staffers in the year ending July, 2001, and opened a 600,000-square-foot headquarters in a refurbished 19th century mental-health facility.

Even inside Sun, though, McNealy was repeatedly warned about the company's "soft underbelly." Three insiders recall a planning session in 1997, when several engineers made a presentation about the increasing power of low-cost chips from Intel. Gene Banman, an exec who had just returned from running Sun's business in Japan, argued that Sun could get a chip from Intel for 30% less than it cost Sun to make an equally powerful SPARC chip. Chet Silvestri, who ran chip design, shot back that his staff would never let that happen. After a 20-minute debate, McNealy put an end to the meeting. "I don't see the problem here," he said, according to one insider. Then he laid down his orders: For the time being, no Sun computers would have Intel inside. Today, Intel's processors are twice as fast as SPARC chips, and McNealy admits that his biggest regret is "not putting Solaris on [Intel's chips] six or seven years ago."

Instead, McNealy continued to insist that Sun rely on its own innovation. A classic example was the September, 2000, purchase of Cobalt Networks Inc. By this time many Sun customers were asking Sun for cheaper servers based on Intel chips and Linux, the increasingly popular open-source software. Zander persuaded Sun's board to O.K. the $2 billion acquisition of Cobalt, which had figured out how to make hefty 51% margins selling specialized "Lintel" servers costing just $1,500. With the help of Sun's 8,000 salespeople, Zander felt he could turbocharge Cobalt's sales and learn how to compete with commodity products.

It never happened. From the beginning, Cobalt was attacked by the heads of Sun's old-line server businesses, who didn't want Sun to invest in the very technology that was threatening their livelihoods. Stephen W. DeWitt, Cobalt's boyish chief executive, became known derisively as the "$2 billion blond." When it came time to set budgets, Cobalt's was cut repeatedly, crippling product development. "The work didn't get done," says former sales chief Jabbar. "The company religion didn't allow it." McNealy says the acquisition was a mistake. "There were some other people who thought it was a good idea, and I trusted them," he says.

Cost-cutting bordered on blasphemy, too. In late 2000, some of McNealy's top lieutenants began to suggest cutbacks, given that the stock market had been tumbling since March, and cash-starved dot-coms were imploding. Even Joy, a devotee of R&D, says he told McNealy to make a bold move. "I thought we should have cut more sooner," says Joy. "But I could never get the discussion to go anywhere." Around New Year's Eve, Zander grew concerned as well. In one meeting with his staff, he said that Cisco Systems Inc. (CSCO ) CEO John T. Chambers had told him Cisco was seeing a large drop in revenues. "Shouldn't we be equally concerned?" he asked the group. "That's when the serious discussions began," says Hambly, the former executive vice-president.

They got a lot more serious on Mar. 8, 2001, when Cisco rocked Silicon Valley by laying off 18% of its workforce. Quickly, Zander's staffers set down the broad outlines of a plan. The consensus was that Sun should cut as many as 20% of its employees, more than 10,000. Many also wanted to jettison the underperforming storage unit and give up on McNealy's hope of creating a viable alternative to Microsoft's Office. "When times are hard, you've got to shoot activities that aren't making money," says Shoemaker, the former server chief.


McNealy would have none of it. Time and again, Zander would end staff meetings by saying: "Well, what do you guys want to do?" says a source who was in the room. As the execs went around the table expressing their views, it was clear everyone believed in some sort of austerity plan -- except McNealy.

McNealy was convinced that the Internet had fundamentally changed the nature of the economy. He thought the highs and lows of the business cycle would be far more extreme and short-lived than in the past, with sharp spikes up and down. "We don't have rolling waves," he said during a conference call with analysts in 2001. "We seem to have real edges." If he was wrong, he promised during one session with his execs, he would "take the heat" from Wall Street and Sun's board. Today, McNealy declines to discuss those gatherings in detail but says: "There's a lot of revisionist history out there."

The executive-suite standoff didn't just cost Sun precious time -- it contributed to a damaging exodus. That had started the previous September, say sources, when McNealy and Zander met for their yearly talk about their personal plans. McNealy had been letting Zander run the show, leading the second-in-command to believe he might soon get the top job. But McNealy said he planned to remain CEO for a few more years -- prompting Zander to decide to move on, ultimately to the CEO post at Motorola Inc. (MOT ). By mid-2002, half of Sun's top-level executives had departed.

As Sun's fortunes waned, customers and insiders, including outspoken software chief Jonathan I. Schwartz, urged McNealy to change course. A key step: McNealy needed to drop his long-running, and distracting, feud with Microsoft. Finally, last June, McNealy says the customer requests won him over. He swallowed his pride and called Microsoft CEO Steven A. Ballmer.


Bringing tech's longest-running feud to a close was not without its funny moments. Microsoft Chairman William H. Gates III visited Sun's headquarters the week of July 4, when the office was closed. The famous exec escaped notice, until he and Sun's chief strategist, Mark Tolliver, went for a bathroom break and ran into a Sun staffer who had come to the office, dog in tow. "He looked Gates straight in the eye, and he had the most stunned look on his face," recalls Tolliver. Still, the meeting stayed secret as negotiations continued over the next few months. Then, early on the morning of Apr. 2, McNealy and Ballmer announced the blockbuster deal. Sun would get $1.95 billion in exchange for calling off two landmark lawsuits. And thanks to a 10-year technology pact, Sun's servers will be certified to run Windows.

The détente was a public-relations coup, at least temporarily. Sun's shares jumped 19% during the two days that followed. But internally, reaction to the day's news was mixed, say insiders. Not only had Sun racked up a net loss of $750 million for the spring quarter, but McNealy had named Schwartz to be Sun's new day-to-day chief. Many say that while he is known for his razor-sharp mind, the ponytailed 38-year-old lacks experience and alienates employees and customers with his arrogant style. When McNealy told other executives about Schwartz' promotion the night before the announcement, the silence on the phone was deafening, according to former execs.

Today, McNealy is as combative and optimistic as ever. During a series of interviews, he sketched out bold plans to jump back to the fore of the server market. Servers based on Sun's throughput computing chips are expected to hit the market in 2006. Meantime, Sun's experiment with freebie low-end servers started in June. The idea is that the low-end servers will open the door for Sun to sell more lucrative software. The company is selling a package of its e-business software called Java Enterprise System to corporations for $100 per employee per year -- a simpler and cheaper alternative to buying from suppliers such as BEA Systems Inc. (BEAS ) or IBM.

Still, McNealy faces a hard slog to win back credibility among customers. For years corporate buyers bought from Sun in part because it seemed to know where tech was heading. Now, many believe Sun spent several crucial years with its head in the sand. "They've always had lots of great things on paper. But when it comes to execution, they're lacking," says Gary Feierstein, vice-president for information technology at Premier Inc., a San Diego company that manages 1,500 hospitals. "They always seem to be behind where they need to be."

Therein lies the true tragedy in Sun's decline. Throughout the company's history, McNealy's risk-taking spirit resulted in innovative alternatives to the offerings of giants such as Microsoft and IBM. Unless he pulls off a longshot turnaround, it may ultimately be a blow to Silicon Valley and even America's technological prowess.

By Jim Kerstetter and Peter Burrows

Before it's here, it's on the Bloomberg Terminal.