OK, so you're the richest company in the world. You have a monopoly in PC operating systems that you've leveraged into overwhelming dominance in desktop productivity applications and programming languages. You have a decent share of the server market and are the third- or fourth-largest commercial presence on the Internet. Beauty, right?
But then the government rewards you for bringing hundreds of billions of dollars into the U.S. economy by trying to split you in two, which lops a third off your stock's value and sets off a string of me-too lawsuits against you from here to Brussels. That's bad, but you'll survive it. Your real problem is that your market space is in the process of moving to cyberspace-without you.
What do you do? Get busy building another "operating system" to dominate your market's new home, of course. You stick with what works until it quits working.
Microsoft's new Microsoft .NET (pronounced "dot net") cyber programming paradigm is all about creating an Internet that looks as much as possible like the desktop space Microsoft has dominated for the past quarter-century. It's a large collection of programming initiatives, sales channel realignments and new Web standards-too large, frankly, for we mortals to grasp. To succeed, Microsoft needs the cooperation of a lot of different audiences-program developers, hardware developers, chip-makers, Web host providers and even competitors-just like back in the day, when it was getting Windows off the ground. This time, it needs its standards to be adopted not just in computing, but throughout the vast electronics marketplace. It will all probably take years to come to fruition (although Microsoft predicts it will take only two years).
During that time, .NET's code base and its associated acronyms, sub-acronyms and marketing catch phrases will morph in ways that not even Microsoft can foresee. Battles will be fought with entrenched interests over the most trivial aspects of each standard or protocol, and those outcomes will determine the ultimate size and shape of .NET. Microsoft will lose some, but win most. Anyway, the details are far less important to Redmond than the ultimate goal: that there be just a dot's difference between the Net and .NET.
"It doesn't matter what standards you're talking about," says Rob Enderle, vice president of the Giga Information Group, a Santa Clara, California, e-business advisor. "They want to own the critical ones between you and whatever it is you are doing on the Web, and it really doesn't matter what they turn out to be."
Microsoft CEO Steve Ballmer has stated that Microsoft is not out to create an Internet monopoly. But, hey, that's what founder Bill Gates said about Release 1.0 of Windows. Unlike government antitrust lawyers, though, analysts don't see Microsoft as an invincible monopolist in a static competitive landscape bullying Sun Microsystems, Oracle, AOL and other Fortune 500 unfortunates with impunity. It's survival time for a company whose desktop franchise is about to run out, says Arthur Williams, a director of ASP research at Giga Information Group. And .NET is nothing less than a bet-the-company attempt to move that franchise to cyberspace-rather late in the game.
"Microsoft was surprised at how rapidly the World Wide Web and Web-based applications have become important to businesses," agrees Dan Kusnetzky, vice president of system software research at IDC. "When people can access them using a browser, it means that the operating system on the client or server no longer matters."
Mike Hogan, Entrepreneur's technology editor, can be contacted at mikehoganentrepreneur@juno.com.
The Shift To Serviceware
Nobody really saw the Internet coming, but Microsoft was later than most to acknowledge the shift in the focus of computing-at least, publicly. In fact, it spent the Web's early years trying to fend off efforts by competitors like Oracle chairman Larry Ellison and Sun chairman Scott McNealy to make "the network the computer."
Now, the plan is to mimic the success of its popular desktop software by creating a new generation of products to be delivered solely over the Web, and to extend that functionality to new Web-enabled devices like handheld computers, PDAs, television sets and maybe a kitchen appliance or two-being careful not to interrupt revenue streams for desktop software.
Microsoft plans to follow up Windows Millennium with Windows.NET, and the next revision of its Microsoft Office desktop suite with Office.NET early next year. Before then, Redmond will roll out half-a-dozen special-purpose Windows 2000 servers for e-commerce, transaction processing, database maintenance, what have you.
For its Visual Basic developers, Microsoft has begun previewing its new full-featured Web development environment, Visual Studio.NET. Apparently, it won't include the Java tools Microsoft built under license from Sun, but rather Microsoft's new C# (pronounced "C sharp") programming language-a Java antidote.
The ultimate goal is to tie all computing devices to .NET, allowing .NET users to access their e-mail, calendar and important files regardless of what device they're using. That's a far taller order than tying various desktop computing applications together under Windows. The operating system, interface, data processing, programming and ergonomic differences between computers of all sizes and would-be Internet devices like phones, PDAs, home appliances and even wristwatches are too numerous to detail here. Suffice to say, it all requires many lines of programming glue, admits Microsoft.NET development group manager Barry Goffe.
The most important element, he says, will be the reprogramming of the Web to allow highly integrated Web services using the new Extensible Markup Language (XML) and Simple Object Access Protocol (SOAP) standards. These are the stand-ards that Microsoft needs the world to adopt-in addition, of course, it needs non-Windows competitors to buy into whatever mezzanine code is needed between their OSes and XML.
Microsoft's first multiyear pro-gramming challenge is to Web-enable its own desktop arsenal and make all of Office. NET accessible in ways only poorly implemented in Exchange e-mail servers today, says Williams. Then, individuals or workgroups will have the option to rent (as well as buy) software from Microsoft and/or ASPs, accessing "programs" in real time through their Web browsers. Microsoft will rent software and other services to consumers through its MSN Web sites and to small businesses through its bCentral.com.
While pricing details are far from finalized, businesses will probably pay a per-user monthly fee for each service used, says Goffe. ASPs will provide the first line of support, but you will be able to get escalated to Microsoft support if necessary, adds Satya Nadella, vice president of Microsoft bCentral.com.
Instead of the historic trend to amass ever-more functionality into a large software bundle like Microsoft Office or Exchange, you'll be able to rent only the services you need, says Goffe. For example, you might buy a copy of Microsoft Office for your desktop PC, but rent only the phone book and calendar functions of Office.NET from an ASP for your PDA.
Not Going Quietly
It may sound like Microsoft is capitulating to the network-as-computer crowd. But look closer at the resources required to fulfill the .NET vision, says Kusnetzky. They won't fit the downsized network appliance that Internet first-movers like Ellison and McNealy see running write-once-run-anywhere applets independent of OS. "It's not clear that people will want to have Windows 2000 and 1GB of disk and 64MB or more of memory on their handheld devices," Kusnetzky jokes.
Actually, it won't be quite that bad. True, Microsoft plans to have a version of Windows.NET on every platform that can handle it-desktops, servers, Pocket PCs. It's not true that every wristwatch and toaster oven that connects to the Internet will have to run Windows, says Goffe. But, he adds, "there is some software that needs to run on the device that understands XML and Web services."
Whatever you call it, that code will force competitors to wait for Microsoft programming direction and bind them to a product release schedule that trails Microsoft's. And in high-tech sales, being first to market is half the battle, notes Enderle.
Can it work? Impossible to say. Microsoft's last three tries at adapting the Windows success to something other than desktops-namely, to Pocket PCs-have won few converts. Palm OS-based handhelds still own something like 80 percent of the market-in part, because Windows and its appurtenances are too cumbersome.
On the other hand, Micro-soft is already pretty rich in Net assets. Internet Explorer, the marketing tactics of which got Microsoft into so much trouble with the Justice Department, is now the cyberspace vehicle of choice 86.1 percent of the time, according to Web analysis firm WebSideStory Inc. Microsoft's MSN group of consumer sites are the third-most-often-visited pages on the Web, according to PC Data Online.
Programming expertise? No one else could even attempt this gambit, points out Enderle, and the market does like someone other than the government to set standards. As for communal standards committees, Microsoft knows how to play them, says Enderle.
But Enderle, Williams and Kusnetzky all see Microsoft's revenue model and margins coming under attack as it migrates away from the usual strategy of selling software licenses to every desktop on a LAN. On the Web, it will have to compete with companies like Oracle that sell limited-use permits where, say, a 300-person company need only buy a license for the 40 or 50 people using the Web service or software at any one time.
Large ASPs like AT&T are more powerful than any of the hardware OEMs Microsoft has dealt with in the past and, because of the antitrust troubles, the prices they get for Microsoft code will be enjoyed by lesser ASPs as well, predicts Enderle. Then there is the acknowledgment that its desktop software is history. "For a company to publicly obsolete every product on the market two years before it's going to ship their replacements-we have seen a lot of companies go under for a lot less," declares Enderle.
But Microsoft isn't just any company, and a preannouncement that might kill another company's product line is one of Microsoft's most effective marketing weapons. "No other company can freeze the market for years while it builds its products," says Kusnetzky. "Even though other people have competing products today that do what Microsoft is talking about, they don't get a full hearing because people are waiting to see what Microsoft will do. It's an amazing example of marketing power."
Williams describes it as a calculated balancing act: "They are shooting their own golden goose in as sensible a way as possible so as to still retain control and mindshare." It will be interesting to see whether Microsoft falls or catapults to new heights.
This story appears in the November 2000 issue of Entrepreneur. Subscribe »
