Microsoft offers a full ecosystem of software to would-be buyers, but its greatest success may actually result from its strategy to present customers with an "and" decision when initially purchasing Microsoft technology, rather than a difficult "or" decision.
Microsoft technology has generally been light enough--in terms of cost and in terms of resources required to deploy and manage it--to enable enterprises to add its technology into existing environments, rather than forcing immediate, wholesale rip-and-replace decisions. With a beachhead in place, enterprises slowly (or quickly) find themselves deploying more and more Microsoft software.
Red Hat does things differently. Today, Red Hat's primary mindset is "replace," as in "wholesale swap out expensive Unix for value-driven Linux." Red Hat recently told me that it's seeing more and more Oracle (BEA) WebLogic and IBM WebSphere customers trading for JBoss at the platform level: rip and replace. This is good, but the mentality is still "or," while most customers want "and." It's about choice.
To gain serious scale Red Hat needs to be more "and" than "or," as it is not big enough to compel wholesale rip-and-replace decisions (which are somewhat rare, anyway) once it starts having to compete with Windows, which is coming soon, and in a contest which it is often likely to lose. Consider that, according to IDC, Microsoft wins more of the Unix replacement business than Red Hat (or Linux, more generally), does. Why? Because Microsoft offers Unix shops a more complete solution, one that goes beyond "or" to include "and" with great developer tools and relatively easy-to-use software.
Integration with the existing (proprietary) world and expansion of its open-source offerings is the model for Red Hat expansion. JBoss works just fine with Windows, and it should continue to do so. Red Hat has partnered with Likewise to enable Red Hat Enterprise Linux to be used with Microsoft's Active Directory, and should continue to do so.
In fact, Red Hat needs to seek more ways to integrate itself into and hence infiltrate the Windows world. Red Hat can't grow like it needs to by demanding that the world beat a path to its open source-only solutions. It needs to give would-be buyers a near-term "and" decision by integrating with customers' existing environments and offering them incremental, open-source alternatives to their existing technology investments. Through things like its RHX program, Red Hat can offer enterprises a way to move to full open-source solutions at their own pace, not Red Hat's.
It's the strategy that Novell has been taking, and it's paying off for Novell. Red Hat should emulate and extend this strategy, not with patent deals and the like, but rather with technology that enables real integration with a customer's existing proprietary environment, then gradually expanding the Red Hat partner ecosystem, easily consumed through Red Hat, as well as adding to its own product portfolio.
Give customers real choice, in other words. "And" is a more palatable choice than "or."