From MSWord to MSWorld:
How Microsoft is Building a Global Monopoly

The Core: Leveraging the Operating System for Desktop Dominance

The operating system, MS-DOS and Windows, are the software core of almost all computers-linking keyboards, the central processing unit, memory chips and all other software together in a functioning whole. While Microsoft has expanded its operations in the last few years, it was through control and sale of desktop operating systems in the consumer market that Microsoft made the fortune from which all other ventures have sprung and it is the operating system that has been the key strategic point of control that has given Microsoft a monopolistic advantage in other ventures. Microsoft is now using many of the same tactics that it used to monopolize the consumer market for the desktop computers to win domination of the business-level and Internet markets.


Microsoft's MS-DOS operating system, begun in controversy and accusations of deceptive business practices, has been the subject of a Justice Department defense decree, and to this day is still the object of residual lawsuits from erstwhile competitors. When IBM licensed the operating system from Microsoft in 1980, the fortune of the company was made as it went on to resell the operating system to almost every company seeking to build computers that were compatible with the new IBM standard. And as IBM was to ruefully discover, in the new age of the desktop, control of the operating system was the true position of economic power in the computer economy.

Like many of its products, Microsoft did not create MS-DOS but bought it from a company, in this case one called Seattle Computer. In the nascent desktop computer market, Seattle's product had no market share compared to the dominant CP/M operating system produced by a company called Digital Research. In fact, Digital Research would continue to contend that Seattle Computer had illegally stolen source code to create their stripped down version of CP/M. But with IBM backing not only MS-DOS but making sure all programs initially shipped for the new IBM PC were compatible only with DOS and not Digital Research's CP/M, MS-DOS quickly came to dominate the market, despite the fact that almost everyone in the industry considered CP/M a far superior piece of software. Microsoft had already mastered the art of leveraging hardware deals, the bundling of applications software and operating systems in a mutually reinforcing strategy to knock off rivals, even when those rivals had superior products.[7]

But Digital Research would continue to try to compete with Microsoft's MS-DOS in the form of advanced versions of CP/M called DR DOS. In fact, many people accused Microsoft of doing little to improve MS-DOS for years after creating its version 3.3 in 1986 until a far superior DR-DOS version was created. In 1990, with no version of DOS to ship in competition with DR-DOS, Microsoft began issuing press releases about the imminent "new DOS" that would include all the features buyers liked in DR DOS and more. Yet Microsoft would fail to ship MS-DOS 5.0 for over a year (slowing DR DOS sales in the meantime) and when it did, it used a new system of deals with hardware resellers to essentially destroy the market for DR-DOS.[8] Microsoft began requiring all manufacturers installing MS-DOS on any machine to pay a license fee to Microsoft for every machine they sold - whether they installed MS-DOS or not. Essentially, the hardware sellers had to pay for MS-DOS in any case so they refused to pay extra for a competing operating system; DR DOS peaked at $31 million in sales in the year DOS 5.0 was introduced and would rapidly be driven from the market.[9] The fact that early versions of Windows issued false error messages that made DR DOS appear incompatible with Windows just added to the destruction of competition for DOS.[10]

The Justice Department belatedly began investigating Microsoft for anti-trust violations and forced the company in late 1994 to sign a consent decree promising not to commit the same offenses in the future, but by this point the damage had already been done and Microsoft owned essentially 100% of the operating system market for all non-Apple desktop computers. District Judge Stanley Sporkin rejected the consent decree for concentrating only on issues that were essentially moot because of Microsoft's absolute victory in the DOS marketplace, while failing to deal with all the emerging controversies over Microsoft's use of its operating systems to support its applications business and other ventures. "It is clear to this court," Sporkin wrote in an impassioned, 45-page decision, "that if it signs the decree presented to it, the message will be that Microsoft is so powerful that neither the market nor the government is capable of dealing with all of its monopolistic practices."[11] However, Sporkin was overturned on appeal and Microsoft was allowed to continue a range of anticompetitive practices based on its control of DOS and Windows operating systems.


Throughout the 1980s, Microsoft had tried to sell versions of Windows to give DOS machines a more Mac-like graphical feel, but it was in 1989 that Microsoft would see Windows as not only the key to leveraging itself into dominance of software applications but as a way to take complete control of the desktop away from IBM itself. Microsoft had been developing IBM's new graphical operating system, OS/2 (which would have replaced Microsoft's cash-cow DOS as well), but in 1990 Microsoft withdrew from that work in favor of concentrating on its release of Windows 3.0 the same year- a relatively amicable separation from IBM that would lead, through IBM missteps, to Microsoft domination.

Part of the problem was that OS/2 was aimed at higher-power machines not generally available, while Microsoft still understood better than IBM how to leverage developer support and third-party software into acceptance of an operating system. And Microsoft was not loath for a little hardball even against former ally IBM, threatening at one point to pull Microsoft's sponsorship of an industry exposition if IBM's top OS/2 executive was allowed to speak.[12]

As for competition with those like Quarterdeck's Desqview which were, like Microsoft, adding a graphical interface on top of DOS, Microsoft was accused across the industry of using unpublicized knowledge of the workings of both DOS and Windows to undermine competitors. Quarterdeck claimed that DOS 5.0 managed network drivers in ways that competitors found impossible to do given published documentation of DOS. As for building an alternative interface on top of Windows, Desqview developers complained that Microsoft products were making "calls" on the operating system that were nowhere documented for competitors to utilize as well. A 1992 book called "Undocumented Windows" by Andrew Schulman analyzed an array of undocumented Windows features that the author discovered by taking the software code apart himself.[13]

Other software companies complained that successive versions of Windows were swallowing up whole categories of software utilities that had once supplemented DOS. By bundling such utilities with the operating system or with Windows, Microsoft could essentially destroy software utility rivals as Microsoft's versions came bundled "for free" with Windows; Microsoft was able to lock-in loyalty to Windows through its array of software utilities and, through the breadth of its operating system licensing, recoup the costs of bundling those utilities with the operating system.

As we will return to when we discuss the conflict over control of the Internet, many analysts worry that the bundling of Internet browsers and other Internet-related software with proposed new versions of Windows will just continue this anti-competitive tool bundling practice. Combined with new acquired technologies like speech recognition software (through the 1997 purchase of Lernout & Hauspie) and handwriting recognition (through the 1996 purchase of Aha), whole classes of software competition will be liquidated as Windows bundled-utilities destroy those software markets with customers unwilling to pay additional money for an alternative when the Microsoft version had to be purchased already as part of the operating system.

Using Windows to Leverage Applications Dominance

The transition to Windows also became an opportunity to steal a jump on rival software companies in key application software, from word processors to spreadsheets to databases. Microsoft application developers were ready to release Windows versions of both its Word word-processing software and its Excel spreadsheets far ahead of rivals, and by the time Wordperfect for Windows or Lotus 1-2-3 for Windows got to market, in the words of Time magazine, "Microsoft already owned the markets."[14]

While this was not instantaneous, the leveraging of the Windows operating system gave it a massive anti-competitive leg-up over rivals, and Microsoft's willingness to spend profits derived from its operating system monopoly to fund entry into new software markets only compounded its advantages. When Microsoft had problems developing its own database software in the early 90s, it readily dropped $175 million to acquire top-rated Fox Software and its team of developers who had created the FoxPro database. In combination with the move to Windows, this new talent gave Microsoft the chance to aggressively enter the database market when its own software team seemed unable to do it on its own.[15]

However, the coup de grace by Microsoft was its promotion of "suites" of software bundled together. By linking spreadsheets, word processors, and databases in one package that could easily exchange data, Microsoft's Office suite was able to overwhelm the makers of individual programs. A key to making such software suites are systems that allow data exchange. Microsoft promoted a proprietary approach called OLE against alternatives like Open Doc which were built around open software standards. While Open Doc was considered an easier to use and more robust approach, fears that future Windows upgrades might undermine any approach other than OLE forced almost all software makers to deal with Microsoft's own proprietary system. And since Microsoft provided little documentation for assuring compatible links between different software packages, customers naturally chose Microsoft suites as the best way to assure their software would work together.[16]

Most competing software had to sell out to larger companies desperately trying to compete with the Microsoft juggernaut. Previously dominant wordprocessor Wordperfect along with Borland's Quattro Pro spreadsheet were first bought by Novell in 1994, then sold two years later to Corel, once only a graphics software company. Borland's Paradox database had once dominated the desktop market but Microsoft's suite strategy had decimated its market share, so Corel in 1996 took over development and marketing of Paradox as well within its suite competitor. Once dominant spreadsheet maker Lotus was so battered by the competition from Microsoft that it was forced to sell out to IBM.

While the massive consolidation of software into competing suites led to a small comeback in sales for Corel and Lotus, by 1997 Microsoft was making an estimated 85% of unit sales of suite software and 90% of all revenues in this key area of home and office software.[17] Between its operating system franchise and its Office revenues, Microsoft had both the dominance of the desktop and the financial war chest to move on to its assault on the corporate computing market and then onto taking control of the Internet.

Next: Moving On Up: From the Desktop to Control of Corporate Computing

< Table of Contents >< Conclusions and Recommendations >< >