Toppling the King: Competitors Hope to Dethrone Microsoft's Monopoly
The business struggle to define the rest of this century's computing -- and Microsoft's primacy within it -- takes on a very new complexion this month. Netscape, the Great Lite Hope, is releasing the beta version of its Communicator, an all-encompassing operating environment that a landslide majority of business software companies is praying will make Microsoft Windows obsolete and revive the overall industry from its death spiral. While the software industry, more than most, requires constant buyer churning of product to keep it healthy, it's been a mere 18 months since the introduction of the previous platform revolution that forced people to reexamine the kind of equipment they buy and to replace all their software. In normal times, the Windows 95 introduction would have jump-started the industry enough to keep most software and hardware companies happy. So why the seemingly premature matriculation of Netscape's newer platform?Two reasons make Microsoft's model open to attack. First is Microsoft's much-resented functional monopoly over business software (and near monopoly over industry-wide profits from selling it). Second, and more important, is the continuing failure of the entire computer industry and its business consumers to address the failure of technology to enhance business success. Tens of billions of dollars into the spending frenzy known as the Computer Revolution, American businesses have still to see even a tiny boost in white-collar productivity resulting from all that capital investment. Studies by Peat, Marwick and the Brookings Institute confirm this failure. American business is starting to fear it is trapped in a cycle of technological bulimia -- devouring equipment and software, then learning that the remaining technogenic problems can only be solved by purging the new technology and gobbling down newer technology. Most organizations have never succeeded in even defining the problems the technology is supposed to solve. On top of this is buyers' realization that Windows (in all its flavors, from 95 to NT) is not significantly more usable and productive than prior versions. Then there is the stark realization that the software industry, aside from Microsoft, is in decline. Low prices for business software may be a boon to consumers, but they provide little or no margin to the publisher. With the extinction of the mom-and-pop and small regional chain software shops, the oligopoly of big national chains keeps wholesale prices low and blocks market access to all but a small handful of well-funded computer giants that can come up with shockingly high shelf-stocking charges.Retail itself is under stress and ready for big changes -- witness the just-announced sell-off of Incredible Universe, the mega of the megastores. Even though the whole system of computer retail has evolved to benefit operations such as Incredible Universe, it doesn't generate enough profits for them to thrive. So business users are upset about the low benefits of current computing, and the industry (except Microsoft) is upset at being blocked from profitability by the current situation. The opportunity is ripe to upset the current equilibrium. The same companies and industry pundits who created the Computer Revolution are now proposing a solution. The competitors they overcame to sell that wasteful vision are also proposing a solution. And they're the same solution. Hold on to your wallets: The proposed solution is to throw away the new technology and replace it with newer technology wrapped up as the Network Revolution. The computer domain's big annual show, Fall Comdex, sets the tone for what passes as thinking in computing's coming year. The show, held in Las Vegas in November, was packed as always with vendors and anxious buyers looking for an excuse to gorge. But, as has become common in the industry, the defining products and concepts have more to do with the manufacturers' business strategy than they do either with technology or with any advantage a user might get from the products. It's customer-orientation turned on its ear: selling what can provide the easiest profits instead of what might have some value. From what I could observe there, people just weren't excited. The two big "trends" were 'palm-top' computers built around Microsoft's new Windows CE operating system, and network computers (NCs, to the syllabically-challenged). CE is supposed to be hot because it looks like Windows 95, and in some respects, it does look like Win 95. But the feature set you get is about even with the electronic organizers that have been available for about six years (like the Sharp Wizard). The downsized versions of Microsoft Word and Excel that run on these systems are those products in name only -- they're bare bones, although they work well. "There are serious shortcomings in the CE palmtops' keyboards," according to Information Week Senior Editor Stuart J. Johnston. Manufacturers have a challenge in shrinking a computer. Without something like IBM's butterfly keyboard design, the size of the keyboard can only be as big as that of the device. "They should license IBM's butterfly design," Johnston adds, "because the way these devices work now, the fastest way to type is to use the eraser end of a pencil to type in one character at a time." Johnston believes the CE platform is an interesting start regardless, but will probably have to wait for handwriting recognition (a technology that has the user write on the screen using a stylus which the computer then interprets) to be fully useful. NCs are supposed to be hot because they'll save corporations money. Like CE, though, this is an old, recycled idea that the market didn't embrace the first time around. NCs are basically intelligent terminals. In the description of the "standard" for NCs, it tells about a device that has no hard disk in the computer, no floppy disk to allow a user to bring her own files, and a powerful processor that runs an environment built around Java, a somewhat developed language that's not standardized completely. What a system built around NCs does is use powerful server computers that store the programs and your files and connect you to it through network hardware. You get to run whatever central administration (the Management Information Services or MIS department) allows you to, and if the server fails (infrequent but irritating) or the network hardware fails (frequent and irritating) or the network operating system hiccups (omnipresent and irritating), you're shut down, frozen out from your programs and files. This directly parallels the model the desktop computers replaced -- mainframe and mini-computers. Why would this model make a comeback? Why would a four-way alliance between software companies, big business buyers, industry analysts, and the press all come together in support of the concept? "The one thing all the vendors involved in NCs have in common is they all hate (and hate is not too strong a word) Microsoft and Microsoft's cabal with Intel," Johnston says. "So the resulting products are distinguished more by what they aren't rather than by what they are." What they aren't is a platform where Microsoft has any advantage, so software built for it can compete in a freer market. That describes the software publisher support. But what about everyone else? The analysts and the press both need eternal change to have something to report about, and preferably something that they can frame as a "war" or "contest." The press needs advertising revenue, and even Microsoft's generous ad-spending habits are not deep enough to float the trade publishers all by themselves. Analysts need to explain things to people in exchange for money, and the money doesn't come if people aren't confronted with new, confusing standards and products. And neither group is responsible for supporting the companies and individuals once they buy the products, so they only understand it from a conceptual level. And the MIS and other people responsible for buying, installing, and supporting technology in big organizations -- what's their incentive? MIS has a pair of incentives. First, it never recovered from the Reformation, the loss of absolute power that slipped away when people started using their own computers instead of MIS'. MIS budgets became harder to justify as computer applications migrated all over the place, run by people with no formal training. Not only was their budget affected, but their priesthood was undermined, too. Any accountant or cost analyst with a little knack for computing could do things quicker, cheaper, and more customized than MIS could for them. MIS desperately needs a Counter-Reformation to try to put the computing universe back in harmony. Another incentive is the dysfunctional one that's been driving American business computing since IBM started marketing its first personal computer. Business computing services, whether internal MIS groups or external consultants such as accounting firms, actually do better when the computer system is stumbling. It accelerates demand for consultative services, stresses the need for their expertise. If everything worked, if software was designed to work the way it should so people could just use it, if systems were deployed properly -- so networks were available to people who need computers to collaborate on projects, so stand-alone computers were available to people who don't need the counter-productive complexity, overhead, and instability of networks -- it wouldn't matter which vendor's technology people were working on. MIS and most computer enthusiasts don't yet have answers for the failure of all this technology to produce results. It's because they're not asking the right question. Computing (and it doesn't matter whether it's an NC running Java or an Intel computer running Windows) doesn't pay for itself because it's almost always presented in ways that require people and work structures to adapt to it, instead of the technology being used as a tool to make people more effective or to produce work of better quality. Few interest groups in these decisions know or care about these issues. The press and analysts are largely ignorant of it. The corporate technology professionals are unwilling to change their world view that puts them and their machines at the middle of the equation. The computer industry is used to making a bundle of money without paying attention to it. And so each new episode of technological bulimia repeats the same binge feast on empty promises and later purging of failed systems. It's a cycle that, as of now, won't be changed by any of the "hot" trends, all of which ignore consumer-side productivity issues for the advantage of provider-side jockeying for profitable niches. And because of the poisonous need of MIS for chronic technology changes and its dependence on (and support for) vendors to provide a stream of evolving technology, the buying organizations probably won't wise up to the irrelevance of all this posturing. Who'll own the computing platform heading into the 21st Century? Probably Microsoft. They combine monopoly power (and resulting low consumer costs intended to pre-empt potential competitors) with the more appealing model of individually run computing. Arrayed against them is the Counter-Reformation of centralized computing in workplaces with MIS getting to choose product (from a wider range of competitive vendors). Previous Counter-Reformations (1988 being the most extensive) built around "diskless workstations" or Unix operating systems failed mostly because people fought to keep control of the computers on their desktops. It's less a Hobson's Choice than a Sophie's Choice, but I'd have to guess that the average white-collar computer user in a large organization, having survived multiple "downsizings" and "re-organizations" is one tough customer, more politically astute than average. And that survivor is more likely to buy into the idea of more control at the desktop rather than have it bound up in the bowels of MIS. All these factors tend to push the balance in favor of Microsoft's continued dominance.