The dawn of the third millennium brings with it a spectacular shift in alliances among the major IT vendors. It is the renewal, after more than a decade of bitter win-lose confrontation, of win-win collaboration between IBM Corp. and Microsoft Corp. Both companies will benefit as a result. More important, renewed cooperation will bring major benefits to user organizations as well, reducing risks and improving returns on IT infrastructure investments.
Collaboration between IBM and Microsoft throughout the 1980s created the PC revolution. It laid the foundation for the distributed computing era of the early '90s and the spectacular rise of the Internet in the late '90s, even though IBM/Microsoft cooperation ended late in 1990 with the acrimonious split over OS/2 and Windows.
For the entire decade of the '90s, the primary divide in the industry has been between what I have long called the ABM ("Anyone But Microsoft") Coalition led (though rather quietly and behind the scenes) by IBM. The cast of characters in the ABM Coalition has changed over the years. In recent years, Sun and Oracle have been the most visible (and vocal) ABM Coalition members. IBM has always, however, been the major mover behind the scenes. The fundamental win-lose confrontation of IBM and its ABM allies defined the structure of the industry throughout the 1990s.
No more. In a spectacular, but inconspicuous manner, IBM has realigned its competitive posture and shifted to a radically different position--far more favorable toward Microsoft than toward Sun and Oracle. The fundamental reason is that IBM's erstwhile allies pose far greater and more immediate threats to it than Microsoft does, and IBM has caught on.
The Sun Also Rises
The revenue prize where IBM and Microsoft are direct competitors today is insignificant compared with the hundreds of millions, billions even, of dollars at stake where IBM is in direct competition with Sun and Oracle. Just about every penny of revenue booked by Sun or Oracle is money won in direct competition with IBM. Only a small portion of Microsoft's revenue, however, comes from areas in which IBM offers directly competitive products. On the contrary, IBM is one of the largest vendors of Windows systems.
Earlier this year, IBM was finally roused from its post-OS wars trauma to realize that it was not Microsoft but Sun and Oracle that have really been picking its pockets lately. The result: a shift by IBM away from a Sun/Oracle bias toward a much more Microsoft-friendly position. Joining with Microsoft last May in submitting the Simple Object Access Protocol 1.1 (SOAP) specification proposal to the Internet Engineering Task Force was one of the first public manifestations of that shift.
This is, however, the tip of the iceberg. As important as SOAP is likely to turn out to be in its own right (and it will turn out to be important--it extends the use of XML to make it a very simple yet very powerful tool for remote-procedure invocation as well as for the exchange of typed data), the SOAP 1.1 submission is far more important as a symbol of renewed win-win collaboration and cooperation between IBM and Microsoft.
Although the dotcom gold rush of the past few years has helped sell billions of dollars of Sun servers and Oracle databases, the IBM mainframe and Windows networks remain as the real centers of gravity for a very large proportion of corporate IT operations. The rush to the Web has brought issues of enterprise application integration (EAI) to the fore and rallied the old (and flagging) CORBA force behind a renewed push based on Java, Enterprise Java Beans and Java 2 Enterprise Edition. Microsoft, of course, offers an alternative approach with Component Object Model Plus (COM+), BizTalk and so forth.
Cutting Back On JAVA
IBM, until this year, was solidly in the pro-Java anti-Microsoft camp, which has greatly boosted the case for Java versus COM+ with most industry pundits as well as with many customers. It has also, however, greatly boosted Sun's ability to sell its servers and Oracle's ability to sell its databases into IBM accounts and inadvertently boosted the potential complexity of enterprise applications integration for many IBM customers.
The truth of the matter is that the approaches of both the Sun/Oracle axis and Microsoft offer reasonable ways to do the job that nearly all enterprises will need to do over the next few years. However, in an organization where IT revolves around IBM mainframes and Windows networks, introducing Java-oriented middleware, which requires skill sets entirely different from what the organization has already developed, is hardly the best way to reduce complexity. On the contrary, to do so will certainly increase complexity (and risk) across the board, at least in the near term. Yet IBM's former pro-Java posture necessarily encouraged such customers to do exactly that!
Even worse, from IBM's point of view, taking a seat in the pro-Java camp in effect was making IBM into a sales prospector for Sun and Oracle. If an organization buys into the notion of Java-based middleware as the key to its EAI strategy, Sun servers and Oracle database software become much more attractive alternatives to IBM mainframes and DB2. IBM, in herding customers toward a Java-centric middleware strategy, not only increases risks and complexity for many of its customers but also helps its most dangerous competitors get into its accounts in the process. It is small wonder then that IBM has decided to opt for greater cooperation with Microsoft.
IBM will not, of course, abandon Java. IBM will, however, increasingly adopt the Microsoft alternative as an equally viable option, work closely in the future with Microsoft and support IBM customers equally, regardless of which alternative they prefer. Perhaps even more important, IBM will work to help customers do what in fact most customers want to do: mix and match Java-centric and COM+-based components on a case-by-case basis, rather than be forced to make ideological choices forced on them by win-lose confrontations between vendors.
The end result of IBM's shift in attitude will be that user organizations will be able to build EAI infrastructures more flexibly and with lower risk. They will be able to choose the best tools for particular jobs without betting on one alternative to the exclusion of the other. There's no doubt that renewed win-win collaboration and cooperation between IBM and Microsoft will benefit both vendors. The biggest winners, however, will be IBM and Microsoft customers.
William F. Zachmann is a vice president of the Meta Group, an IT research company based in Stamford, Conn.