Can Sun Weather the Dark Days?

Even before its huge loss, CEO McNealy was putting in place a comeback strategy

As profit reports go, this one was a doozy. On Sept. 29, giant Sun Microsystems (SUNW ). Inc. announced that in the first quarter of fiscal 2004, which ended the day before, it would record a net loss of as much as $304 million, compared with the $61 million loss that Wall Street had been expecting. Ouch. The red ink, together with a $1 billion noncash charge to the previous quarter, signals ominous days ahead for the once-heralded innovator.

Boil all the financial issues down, though, and they amount to this: For the last two years, CEO Scott G. McNealy and Sun have had difficulty selling what customers want -- low-cost computing based on open-source technologies that aren't owned by a single company, such as Sun. The company became one of the behemoths of Silicon Valley by selling six-figure computer servers that run Sun-made software on Sun-made chips. But over the past two years, the industry's hot sellers have been $2,000 to $10,000 servers that run on Intel chips and low-cost Linux software made by companies such as Red Hat and SuSE. While Sun shunned Intel and Linux, its market share tumbled, and rivals IBM, Hewlett-Packard, and Dell gained ground. "They just missed it," says analyst Steven Milunovich of Merrill Lynch & Co.

But don't write off McNealy & Co. just yet. Over the past 20 years, Sun has floundered two other times, and the CEO has managed to right the ship. He still has a chance to do it again. The irony of the recent financial hit is that it comes just as McNealy has finally put a potentially winning strategy in place after two years of missteps. In May, Sun introduced a host of competitive, low-cost servers that use Intel chips and Linux software from Red Hat and SuSE. That should help minimize customer defections to low-end rivals. The company also is making progress in its push into software, which carries much higher margins than the hardware business.On Sept. 16, Sun debuted a package of integrated software with simple, flat-rate pricing that could boost revenues and profits.

SPEED AND SIMPLICITY. Longer term, McNealy is continuing to do what he has always done -- swing for the fences. Despite rising criticism, Sun is maintaining its $1.9 billion annual research-and-development budget. That's 18% of Sun's expected revenues this fiscal year, up from 10% in 2001. McNealy wants to spend big on new ideas that could make the company a tech leader again. Sun is working on chip research that could create a hundredfold jump in processing speeds and software that would deliver computing power as simply as electricity. Either of those breakthroughs could brighten Sun's long-term prospects.

No question, McNealy still has much work to do. Sun has lost so much momentum that customers may be reluctant to buy from the company because they fear it won't be around to support its products. In a recent survey, Milunovich found that a third of Sun customers are considering moving to other companies' products. Customers also may question McNealy's long-term commitment to Linux, considering his 11th-hour conversion. And he still needs to get Sun's costs in line. Although Sun eliminated 6,000 of 43,000 jobs over the past two years, it has not cut costs as deeply or aggressively as rivals. As it stands, Sun is expected to lose $302 million for the fiscal year ending in June, 2004, according to Banc of America Securities (BAC ). That's on $10.4 billion in revenues, down from a peak of $18.2 billion in fiscal 2001.

Still, some customers think Sun is back on track. Mark Dickelman, chief technology officer at Bank One (ONE ) Corp. in Chicago, says Sun is taking security and reliability more seriously than its rivals. The result is that Bank One is considering boosting its spending with the company. "They're talking about things that really matter," says Dickelman.

Sun may be getting its strategy together at the right time. While corporations have kept their purse strings tight over the past three years as the economy has sputtered, they're now starting to ramp up spending on tech gear. According to the Commerce Dept., corporate tech expenditures rose 3.2% in the second quarter, the largest increase since spending began to tick up last year.

The strategic repositioning has meant wrenching change for McNealy and Sun. In addition to selling Red Hat and SuSe Linux on Sun-made boxes, McNealy is protecting his flank on the low end by selling Sun's own Solaris operating system on cheap servers running on Intel chips. It's a pride-swallowing move for Sun, which stopped pushing Solaris for Intel in the late 1990s to focus on its own chips. McNealy, not someone who readily eats his words, admits that was a mistake. "We didn't exactly jump on the bandwagon," he said at the product's launch in May.

Making Sun a serious player in software has been no easier. Last year, after years of half-hearted efforts, Sun finally began the hard work of improving its so-called infrastructure software, which acts as the digital foundation for big e-business systems. The strategy made sense, since business software carries gross margins north of 80%, compared with about 30% for most hardware. The execution, however, was terrible. Going into this year, Sun was a distant third to IBM (IBM ) and BEA Systems (BEAS ) in the infrastructure-software market. And Sun, for all its talk of selling an integrated software product, didn't make sure all the pieces worked well together until this summer.

Now all those software pieces are tied together. And Sun is offering a simple, $100-per-employee price that it believes can cut through the confusion in a software industry where pricing can be based on anything from a server's processing speed to the number of people using the software over the Net.

So there may yet be a path out of the woods for Scott McNealy and Sun. McNealy has a solid strategy in place. Now he has to make it work.

By Jim Kerstetter in San Francisco

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