In a slack antitrust environment, Watson swiftly weeded out virtually all of IBM's once many electromechanical equipment rivals.
To its huge credit, IBM and Tom Watson Jr. managed to retain this near-monopoly position through the long shift to totally electronic computing, creating the mainframe colossus of the 1960s and 1970s. Since then, monopoly power has been progressively and unavoidably dissipating. First arrived minicomputers, where IBM could never get more than a third of the business. Then, certainly came the PC and IBM's crucial decision to essentially pass its power on to Microsoft and Intel and, much less directly, to Cisco.
As great as the power of these three companies is today, it's still restricted to a few relatively narrow industry segments that account for only a tiny share of total industry revenue. Further, even the joint global power of the three giants is nowhere near what IBM alone enjoyed in its halcyon days. Understandably, the computer business has become more competitive over time. But now the Internet has come up to finish off the job.
Already, the emergence of non-PC devices and the invariably problematic plug-compatible chip competition have helped confine Intel. Now, the accelerating shift from packaged software to Web services promises to restrain Microsoft's once uncontrolled ambitions. Finally, the shift from in-house corporate networks to public carrier services will inescapably undo Cisco's still-iron grip on the data-communications-equipment industry. All three companies should continue to flourish tremendously, but their power has already begun to fade.
People frequently ask who the next Microsoft or Intel will be. The simple answer is "No one." Although there will certainly be many huge new Internet companies, there probably won't be any major new monopolists. No company will be able to control Internet transmission, content, commerce or access devices, although Microsoft and Intel still have a slim chance to eventually gain control of the Web server business.
The reality is that one of the main reasons IBM, Intel, Microsoft and Cisco have grown so rich is that their customers have been so locked in. Software compatibility has always been at the heart of sustainable IT monopolies. But on the Web, software is subordinate to services, and services are much more difficult to control. Judge Jackson's quandary is how to deal with today's still-powerful Microsoft monopoly at the very moment it gets clear that it should finally fade away.
In this upsurge of movement to network computing, Microsoft stands to lose a great deal. One rationale is that network computer applications are run on powerful midrange computers known as servers. The business world is moving its applications to the server. While the trend is expected to favor Microsoft's Windows NT server operating system, it also augurs ill for the now omnipresent Windows 95 and other desktop-based Windows versions.
That's because the new network computers no longer require the kind of number-crunching power of a high-tech $900 Pentium chip, leave alone the sophisticated word processing, spreadsheet, database, and other applications of the kind that Microsoft bundles