The Uncivil War Inside Microsoft

Why did Gates & Co. seem so hamstrung during the antitrust trial? A new book, Breaking Windows, blames an internal feud between the Windows and Internet camps

By Stephen H. Wildstrom

In the fall of 1996, Microsoft Senior Vice-President Brad Silverberg, who headed the company's Internet Platform & Tools division at the time, confided a stunning bit of news to me. The next version of Microsoft's Visual Basic development package, he said, would be able to create programs for Web sites using both Microsoft's proprietary ActiveX technology and Sun Microsystems' Java. At the time, Java was the hottest buzz in high-tech -- and was widely viewed within Microsoft as a threat to the continued dominance of Windows.

Visual Basic never did learn to speak Java. Like anyone else who followed Microsoft in the mid-'90s, I knew the company was rent by a nasty high-level struggle between Silverberg, who had led development of Windows 95 and the Internet Explorer browser, and fellow Senior Vice-President Jim Allchin, who had expanded his domain from the business-oriented Windows NT to all Windows development. The two executives had sharply conflicting visions for Microsoft: Should the company bet its future on the emerging Internet? Or on its proven, Windows-centric business model?

The Silverberg-Allchin struggle, which dragged on after Silverberg severed his relationship with Microsoft in 1999 and resonates even today, forms the heart of Breaking Windows: How Bill Gates Fumbled the Future of Microsoft (The Free Press, $25) by Wall Street Journal reporter David Bank. This solidly reported and smart book explains how Microsoft went, in a few short years, from a behemoth bent on world domination -- and one likely to have achieved it -- to a troubled, albeit very profitable, company uncertain about its role and future. Along the way, Bank captures the human drama of the struggle.


  Bank details the disillusionment and ultimate departure from the company of many of Microsoft's best minds, Bill Gates's increasing struggle to understand and control what had become a vast enterprise, and finally, his reluctant passing of the reins of active management to new CEO Steve Ballmer.

The Web and its open standards were the future, Microsoft's Internet faction contended. Perhaps, replied the software experts inside the company, but Windows was the wellspring of the company's success and would remain the key to its financial future. The Internet forces had been in the ascendancy since December, 1995, when Gates publicly embraced the Net as the future of the company. But the move turned out to be more of a tactical feint designed to thwart upstart Netscape Communications than a strategic reorientation.

By the time of my visit to the Microsoft campus in Redmond, Wash., in the fall or 1996, Allchin's Windows team was back on top. Within months, the Internet Platform & Tools group had been broken up and, not long after, Silverberg left on an extended sabbatical from which he never really returned.


  The battle for the soul of Microsoft received very little media coverage, probably because it coincided with Microsoft's very public fight with government antitrust lawyers. There's little evidence that the antitrust suit had any significant effect on the Windows vs. Internet struggle, other than distracting top managers. But the internal fight had a dramatic, if largely hidden, impact on the trial.

Bank makes a compelling case that the notorious internal e-mails detailing Microsoft's plans to smash Netscape and dominate the Web -- evidence that did immeasurable damage to the company's position in court -- seem far more benign when seen in the context of the internal war in which they were written. And executives' unwillingness to air the company's dirty linen during the trial helps explain Microsoft's otherwise baffling failure to mount an effective defense.

Allchin "was unable to explain in public his strategic and legal position without exposing the power struggles within Microsoft and the depths of its internal confusion," writes Bank. "At the same time, government prosecutors were loath to explore the meaning of the split between the Windows and the Internet teams. They feared they might undermine the value as evidence of Allchin's impassioned e-mails. This odd and self-serving complicity between defendant and plaintiff prevented one of the most important threads of the Microsoft story from being aired at the trial."


  Bank's prose is workmanlike, but the story he tells more than makes up for any deficiencies of literary style. Like nearly all books published these days, it seems, Breaking Windows is marred by sloppy editing. Characters are introduced twice, sometimes three times, as though they had never appeared before. Legal guru Lawrence Lessig is identified once as a Harvard law professor, and once as from Stanford's law program -- without the explanation that he moved west in 2000. But these are quibbles that barely mar an otherwise excellent effort.

Bank completed the book between February, when the Court of Appeals heard arguments in the case, and its June 28 ruling, which upheld key findings against Microsoft while tossing out Judge Thomas Penfield Jackson's order that the company be broken up. Whatever Microsoft's legal future, it is moving on. The Herculean effort required to ship Windows 2000, which, like the antitrust case, is a subtext of the book, has been accomplished successfully. The October shipment of Windows XP will accomplish Microsoft's long-cherished goal of unifying its "professional" operating systems, based on Windows NT, and its consumer versions, whose legacy stretches back to MS-DOS.

Microsoft is now putting the bulk of its energies into its .Net initiative, an ambitious plan to provide a broad variety of seamlessly integrated consumer and business services over the Web. Faced with America Online's dominance of consumer markets and the big head start other companies, most notably IBM, have in providing Web services for business, Microsoft is back in the position where it has always enjoyed its greatest success, as a scrappy underdog.


  Times have changed, however, since Microsoft came from behind to smash Lotus, WordPerfect, Netscape, and others. The .Net business model depends heavily on companies that want to do business on the Web hooking up with Microsoft, but the company's long history of leaving a trail of battered partners makes this a difficult proposition.

Within Microsoft, the old doubts seem to persist: Is .Net really a breakthrough, taking the company into a new world of open standards and partnerships? Or is it the newest "embrace-and-extend" ploy designed to breathe new life into a flagging Windows franchise, especially on the servers that run Web sites? You can read the evidence either way. On the one hand, the new Microsoft has made peace overtures by offering to let some .Net services run on platforms others than Windows. On the other, the old Microsoft has stirred a great deal of political and legal controversy by tightly binding features in Windows XP to .Net services, possibly to the detriment of competitors.

Bank agrees that .Net is by no means a guaranteed win for Microsoft, but in the end concludes that Chief Software Architect Gates will probably pull it off: "Gates has offered no public unburdening about his private torment as he bore both the weight of expectations and the pain of unrelenting attacks. There has been no catharsis, no sense of what he may have learned....He has acknowledged no mistakes save for his regret that he didn't more effectively tell Microsoft's story.

"So metaphor will have to do. I [Bank] say he shoots the moon."


  As for me, I'm less sure. The lightning that transformed a tiny software startup into the world's most important technology company in 15 years isn't likely to strike twice. And Gates himself seems more comfortable today as a philanthropist than as a technology mogul. The Bill & Melinda Gates Foundation, a huge donor to minority education and Third World health concerns, seems more likely to dominate the second act of Gates' life than Microsoft.

Microsoft will certainly survive and, more likely than not, prosper. But the odds are that it will never again reach the level of either dominance or profitability it enjoyed during the mid-1990s. And for the world at large, that's probably a good thing.

Wildstrom is Technology & You columnist for BusinessWeek

Edited by Douglas Harbrecht

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