The Open Road

Red Hat is at the top of its game right now, delivering quarter after quarter of impressive performance despite (or, perhaps, because of) a global recession. But it wasn't always thus. Despite a meteoric initial public offering in 1999, Red Hat spent years fumbling about for a winning game plan, dabbling in technologies that took it far beyond its core competence in operating systems.

Small wonder, then, that Red Hat today hasn't risen to the bait to move beyond its core competence.

In fact, it may well have been the burden of its IPO that set Red Hat scurrying to marry open-source ideology with hard-headed business acumen, a thought prompted by Red Hat's bizarre history of acquisitions. This history suggests that much of Red Hat's laser-like focus on core infrastructure today may stem from its wild forays into just about everything else in the past.

In 2000, Red Hat made five acquisitions: Cygnus Solutions (at $674 million, the most expensive by far), Bluecurve, Wirespeed Communications, Hell's Kitchen Systems, and C2Net. From secure Web servers (C2Net) to embedded systems (Cygnus) to e-commerce payment processing (Hell's Kitchen) to software for enabling wireless devices to communicate with the Internet and private networks (Wirespeed) to performance management (Bluecurve), Red Hat's acquisitions were scatter-shot and, ultimately, mostly a failure.

True, the acquisitions brought some excellent talent to the company (Michael Tiemann came from Cygnus, Michael Evans came from Bluecurve, etc.), but in terms of products, it's telling that the markets served by these early acquisitions are not markets where Red Hat does much today. Embedded? E-commerce? Nope.

In later years, Red Hat's acquisition strategy initially proved just as chaotic, but as the acquisition pace slackened, Red Hat's focus tightened. So, while 2001 saw Akopia (e-commerce) and Planning Technologies (consulting) join the fold, 2002 brought NOCpulse, a server monitoring solution, which bolstered the still-nascent Red Hat Network that Red Hat launched in September 2000.

Red Hat was no longer casting about for a business model. It had already discovered a winning model with the launch of Red Hat Advanced Server in March 2002, a move that would be augmented just a year later with the release of Red Hat Enterprise Linux in March 2003.

Since that time, Red Hat has eschewed anything that might distract it from its stated central mission: commoditizing and delivering superior value around core IT infrastructure, including (today) the Linux operating system and JBoss middleware.

In its early days, flush with IPO cash, Red Hat could afford to experiment with e-commerce, embedded computing, and applications (e.g., Red Hat actually picked up several content management systems as part of its spending spree). Today, Wall Street is unforgiving and would likely punish speculative acquisitions.

For new open-source companies, there's a lesson in Red Hat's current focus, but also in its early profligacy: good companies learn from mistakes and can overcome their teenage foibles.

There's another lesson: sometimes your competitors can pay dividends. After all, one of Red Hat's early investors was...Novell, an investment that made Novell over $100 million. That's foresight!

Speaking of which, the other company that has far more money than discipline is Google. Google may well be like early Red Hat, except orders of magnitude bigger. Could Google learn something from Red Hat's shift from an early affliction of ADD to solid, disciplined execution? I think so.

Follow me on Twitter @mjasay.