This is the second chapter in a real-time eBook writing project I launched and explained last week. The following is one of a number of stage-setting chapters to follow. Comments, corrections and suggestions gratefully accepted. All Microsoft product names used below are registered trademarks of Microsoft.
Chapter 2 – Products, Innovation and Market Share
Microsoft is the envy of many vendors for the hugely dominant position it enjoys in two key product areas: PC desktop operating systems – the software that enables and controls the core functions of personal computers – and “office productivity software” – the software applications most often utilized by PC users, whether at work or at home, to create documents, slides and spreadsheets and meet other common needs. Microsoft’s 90% plus market share in such fundamental products is almost unprecedented in the technical marketplace, and this monopoly position enables it to charge top dollar for such software. It also makes it easy for Microsoft to sell other products and services to the same customers.Microsoft acquired this enviable position in each case through a combination of luck, single-minded determination, obsessive attention to detail, and a willingness to play the game fast and hard – sometimes hard enough to attract the attention of both Federal and state antitrust regulators. Early on, Bill Gates and his team acquired a reputation for bare-knuckle tactics that they sometimes seemed to wear with brash pride. Eventually, these tactics (as well as tales of Gate’s internal management style) progressed from industry rumors to the stuff of best sellers, like Hard Drive: Bill Gates and the Making of the Microsoft Empire.
With the emergence of the Web, of course, the opportunity for widely sharing stories, both real (of which there were many) and apocryphal, exploded. Soon Web sites such as Say No to Monopolies: Boycott Microsoft enthusiastically collected and posted tales of alleged technological terror and dirty deeds. More staid collections were posted at sites such as the Wikipedia. The increasing tide of litigation involving Microsoft, launched not only by state and federal regulators but by private parties as well, generated embarrassing documents. Such original sources were not only difficult to deny, but almost impossible to repress in the age of the Web – and of peer to peer file sharing as well.
Moreover, while Bill Gates and his co-founders rarely displayed the creative and innovative flair of contemporaries like Apple’s Steve Jobs, neither were they troubled by the type of “not invented here” bias that sometimes led other vendors to pursue unique roads that sometimes led to dead ends.
Instead, Gates and his partners paid careful attention to what was available and successful in the marketplace, and had no compunction about copying the basic designs of products already in the market. In any event, while Microsoft’s developers could be creative at the code level, its success was built far more on innovation in business strategy rather than innovation at the level of product design. In the early days of Microsoft, this path was eminently available, because software could not then be patented. Even after that rule was conclusively overturned in the courts in 1998 (ruling on a patent initially granted in 1993), copying the concepts, but not the actual copyright-protected code or screen displays of existing software remained fair game for a time. Only gradually did a sufficient number of software patents accumulate to begin to inhibit the ability of one vendor to create products that substantially replicated the functionalities of competing products.
And what Microsoft couldn’t copy, it was more than happy to license or buy. In some cases, competitors felt that Microsoft went much farther, and Microsoft soon acquired a reputation for overly aggressive, and worse, tactics of many types. Nor did the vendor discourage that reputation entirely as it developed an increasingly brash, steamrolling corporate persona that suggested the inevitability of its success in any niche that it chose to enter. Sometimes, these tactics landed Microsoft in court. In one of the more notorious examples, a company called STAC claimed that Microsoft first gained access to its disc compression technology under an agreement between the two companies, and then appropriated that technology into its own products. Microsoft outlasted STAC in court, and STAC eventually went out of business. Microsoft steamed forward, stronger than ever.
The strategy of buying rather than building served the company well at the most critical juncture in the company’s history, when Microsoft famously agreed in 1980 to provide an operating system to IBM for the new line of “personal computers” that IBM wanted to introduce into the market. In an effort to catch up with early market entrants like Commodore, Tandy, Altair, and especially Apple, IBM decide to diverge from its historical development practices. Instead of designing and building the IBM PC from the ground up, it decided to use as many third party components as possible – including both the CPU (or central processing unit, the silicon chip that provides the engine that drives the computer) as well as the operating system (OS). For the CPU, it bypassed an advanced, available and suitable chip of its own, and turned instead to Intel, ordering less powerful chips in that vendor’s new 16 bit 8086 chip line. And for the operating system, IBM turned to a largely unknown software company called Microsoft after negotiations with a larger company named Digital Research broke down. In a decision that not only secured the future success of Microsoft but also established the technology landscape for decades to come, IBM famously agreed to license, rather than buy, that OS from Microsoft.
What IBM did not know was that Microsoft didn’t have an OS to sell, and wouldn’t develop one in time to meet the requirements of the contract. After signing the agreement with Big Blue, Gates scrambled to license the original operating system from a nearby hardware company, called Seattle Computer Products, for relicensing to a “secret customer.” SCP had first named that OS “QDOS,” for “quick and dirty operating system,” because it had been developed for internal product testing purposes with only a few man months of effort. Reportedly, Microsoft paid only $25,000 to license QDOS, and then a further $50,000 to buy the OS outright (later, it paid $1 million to settle a suit brought by SCP in connection with the transaction).
Ironically, while IBM succeeded in giving the “PC” name to desktop computers, PCs never became as successful a business line as it had hoped, in part due to these early design decisions. Eventually, IBM sold the business in 2005 to Lenovo, a Chinese company. Nor did SCP, the developer of the OS Microsoft bought, flourish: it’s own 8086 related hardware products did not sell well, and eventually it went out of business. But IBM’s willingness to license rather than buy the OS it used in the IBM PC allowed Microsoft to license the same operating system (which it now called MS-DOS) to other vendors as well. And since IBM had opted to buy Intel chips rather than use chips of its own design, other vendors could buy them as well.
The result was a rapidly growing market for what came to be called “IBM PC clone” computers, which was unfortunate for IBM, but wonderful for the success of the PC market in general, because it supplied a “de facto” standard platform upon which independent software vendors (ISVs) could economically create software to sell to an ever-expanding market of compatible computers, all built by different vendors. The growing availability of useful application software helped make clone PCs more attractive than the proprietary design computers still being sold by Commodore, Tandy, Altair and Apple. Moreover, competition between the new, largely fungible products of the clone vendors naturally drove down the prices for all of their computers (and to prices that IBM sadly could not profitably match). This fierce price competition drove clone sales up, and encouraged more people to found new desktop software application development companies to sell software to run on the victorious “WinTel” platform, since there were more and more customers to buy their wares. The final result, of course, was enormous profits for Microsoft and Intel, and lower and lower profits for everyone else.
Microsoft’s decision to buy rather than build the initial MS-DOS operating system has typified its strategy ever since, underlining the value of innovation in business and marketing strategy over technical innovation. In the years that followed the success of MS-DOS in overwhelming the desktop, Gates in particular played Rome to Steve Jobs’ Greece. Just as the Romans admired, copied, and then expanded upon the innovations of their neighbors to the east in the ancient world, Gates never thought twice about adopting the creations of early market entrants and product visionaries like Jobs. And just as Rome eventually conquered the Greeks and seized its colonies, Microsoft swiftly came to dominate the marketplace for the desktop OS software that the first home computer manufacturers had created.
Nor did Gates stop there. When Apple launched a revolutionary new computer called the Macintosh in January of 1984 with the first commercial implementation of a “graphical user interface,” or GUI, it appeared that Apple might begin to regain market share. Unlike MS-DOS, which required users to navigate solely using cursor keys and to sometimes learn and type obscure commands, Apple’s Macintosh was far easier and intuitive to use, making it particularly popular with schools. What made it so much easier to interact with a Macintosh than a PC clone was not only the GUI, but another technology that Apple was the first to commercialize: the mouse.
Microsoft reacted by launching a GUI development effort of its own, copying essential elements of the Apple GUI in the process. It launched the first version of what it called Windows in November of 1985. Not surprisingly, Apple took Microsoft to court, alleging copyright infringement – but Microsoft pointed to a pre-existing cross license between the companies, claiming it gave it the right to knock off the Macintosh GUI. Although that license had been intended for a different purpose, a judge agreed that it could be read broadly enough to cover the rights Microsoft needed to copy Apple’s innovative GUI as well, and Apple’s fortunes began to recede once again.
Still later, Gates famously underestimated the importance of the Internet, allowing a brash, venture-backed startup called Mosaic Communications Corporation (its name was soon changed to the more familiar Netscape Communications) to both create and soon own the market for software that made it easy to access and “browse” Tim Berners-Lee’s new creation: the World Wide Web. Owning the access point to the Internet and the Web, Netscape (and eventually Microsoft) realized, could permit users to easily access and alter files from any operating system linked together via a network, a capability that might undermine the source of Microsoft’s dominance. Once that realization belatedly struck home, Gates reacted swiftly, making an aggressive entry into the browser market. Soon he succeeded in supplanting Netscape – by largely copying Netscape’s Navigator browser, and then by giving what Microsoft called Internet Explorer away for free. Even more effectively, Microsoft bundled its new browser software with the Windows operating system, so that every new PC that shipped with its Windows operating system (which is to say, virtually every computer not sold by Apple) also included a free copy of Explorer.
Gates had learned another valuable lesson from Microsoft’s success in the PC clone marketplace that he would never forget: the importance of attracting and holding the attention of a wide range of ISVs willing to develop products compatible with the WinTel platform, and thereby make that platform more attractive to potential new customers at little or no cost to Microsoft. As time went on, Microsoft played a sort of cat and mouse game with the growing number of ISVs in the desktop marketplace. On the one hand, Microsoft provided technical information and marketing support to ISVs to provide incentives for them to “port” their products to those of Microsoft, and to upgrade those products as Microsoft upgraded its own to keep the two compatible. But from time to time, Microsoft would also launch new products in competition with those of ISVs, or worse, simply add the functionalities of those products to Windows itself at no additional cost.
When it did, the result was both good and bad for the end user, who got more and more software functionalities and services for the same price. Meanwhile, processor speeds, storage and other capabilities continued to increase as well, making a desktop computer more and more desirable. But at the same time, each time that Microsoft eliminated a competitor, choices and innovation diminished, because the incentive for Microsoft to improve those functionalities abated, and potential new competitors had little hope for success in selling for a fee what Microsoft was already making available for free. In the case of Explorer, for example, Microsoft allowed its rapidly developed browser to languish for almost seven years without a major upgrade, after vanquishing Netscape Navigator – roughly half the entire lifetime of the Web at the time. Only when a resurrected version of Navigator, called Mozilla, began to gain market share did Microsoft issue a new release of Explorer.
After the desktop version of Windows and the Office productivity suite, Microsoft’s next most popular and profitable product line is the version of Windows created for use on servers, the computers that most business PCs are linked to in networks. These computers are more powerful than PCs, and (among other tasks) host application software – the software that business people use to perform discrete tasks, such as creating documents and working with databases. These larger computers “serve” needed application software to the PC’s that individuals use at work, which become their “clients” in the client-server architecture that has become pervasive in most business environments today.
Microsoft’s entry into the server space was both inevitable and timely. Inevitable, because having conquered the desktop, it needed to attack another platform in order to grow its OS sales faster than the desktop market was growing. And timely, because of changes in the marketplace itself.
Those changes involved the evolution of the IT environment away from proprietary, closed systems where a single vendor would sell the entire “stack” of hardware and software to a customer, and then own that locked in customer for many years. In the proprietary world, desktop users sat at “dumb” terminals comprising a monitor and keyboard, but no resident application software, linked to huge “mainframe computers. Later, terminals more commonly connected to the minicomputers that for years were the foundation for the transitory success of companies like Digital Equipment Corporation (DEC), Data General and Wang Laboratories – all now gone. These minicomputers in turn gave way to what we now call servers, and these servers, like many of the minicomputers they replaced, ran variations of a program called UNIX, originally developed by Bell Laboratories. But each vendor used a somewhat different version of UNIX, making it still difficult for customers to switch from one vendor to another.
Microsoft came late in the game to the server marketplace, but its timing was right, in that the next evolutionary transition was beginning to take place in hosted terminal architecture. And while many major companies, such as IBM, HP and Sun were well established in this marketplace, Microsoft had good field position as well for the game that was to follow. This was because most Microsoft’s PC customers already used servers, and these customers could gain technical advantages by buying both server and desktop products from the same vendor so that they would more easily work together. Customers could expect favorable bundled pricing as well, and that pricing could be very attractive indeed as Microsoft first entered the market for server operating system software.
Still, fourteen years after introducing its first version of Windows usable on servers, Microsoft has today not a 90% market share, but a rising [42%] position, sharing the server niche with declining sales of UNIX systems, which continue to be offered in various proprietary flavors by vendors such as IBM, Sun Microsystems and Hewlett-Packard, and also with an “open source” operating system called Linux, which is dominant in applications such as Web hosting, and enjoys a roughly equal market share with Microsoft overall. The Linux operating system is available in multiple distributions (or “distros,” for short) for free from a variety of commercial companies, such as Red Hat and Novell, and from numerous non-profit, volunteer projects as well, with names such as Slackware and Debian. Together, Linux and other open source software (about which much more later) has raised a formidable challenge across the board in all existing and emerging computing environments in which the Microsoft competes – servers, desktops, and mobile devices, such as personal data assistants (PDAs) and mobile phones.
Still, despite Microsoft’s huge early successes, enormous market capitalization, and huge cash reserves (typically over $30 billion), Microsoft has launched far more failed and semi-successful products than dominant ones. Over the years it tried and failed to achieve success in areas such as portable touch screen LCD monitors (SmartDisplay, launched in 2002 and cancelled the following year), and in a rare and ill-fated effort to match Apple’s level of creativity in user-friendly product design, called Bob – an interface intended to make learning how to use a computer more simple. Bob proved not only to be unsuccessful (and some thought even more annoying than Clippy, Bob’s lineal descendant), but was soon cancelled, earning the unfortunate honor of being named “worst product of the decade” by CNET.com. Today, Microsoft has increasing revenues from a variety of newer products and services, such as its Xbox game stations, where it competes with Sony’s PlayStations and Nintendo’s GameCube, Web-based advertising, where it lags Google and Yahoo, Zune portable music players, where Microsoft (and all other competitors) have been hopelessly outsold by Apple’s wildly popular iPod line, and the CE OS for portable devices such as phones, where a variety of Linux distros is rapidly becoming dominant.
Microsoft’s difficulty in entering markets that were already dominated by well-established and capable competitors is doubtless attributable in part to its culture of copying already successful products rather than entering the marketplace with more innovative alternatives. As a result, it has operated many of these new business lines at a significant loss – often in the billions of dollars per year – as it struggles to gain market share. These billions continue to be provided by its two flagship products, operating system and office suite software.
The picture that emerges from even such a cursory review of Microsoft’s history, as well as its current financials, is therefore of great and ongoing dependence on just two product lines – Windows and Office, although that picture is gradually changing over time. In 2005, Windows and Office contributed $[X] billion and $[Y] billion out of $[Z] in total revenues, and a much larger percentage of its operating profits (Microsoft projects that the equivalent numbers for its 2007 fiscal year will be X, Y and Z for 2007). Clearly, under no circumstances can Microsoft afford to see its dominance in either of these product lines be eroded.
Moreover, Microsoft knows only too well how it acquired control of the office productivity suite marketplace to begin with, and how its lock in of the same niche has been sustained.
Have YOU Discovered the Alexandria Project?