by Bernard Golden

What CIOs Don’t Get About Open Source

Jan 29, 20074 mins

I found the comments about my recent posting “Why CIOs Don’t Care About Open Source” pretty interesting. Several of them dismissed the importance of open source because the authors were too busy delivering “business benefit,” which seemed to be wrapped up in implementing packaged apps. Open source was cited as too much RTFM and not enough … business benefit. In other words, because open source isn’t enough like packaged apps from Microsoft and Oracle, there isn’t any point for CIOs to consider using it.

If there was ever a time for CIOs to think of their job as taking the latest release from Gigantic Enterprise Vendor and figuring out how to spend gobs of money in installing, configuring, and integrating it, that day has passed. Today, businesses are being turned inside out by digitized work processes, automated supply chains, ad-hoc coalitions of organizations that come together to aggressively address nascent markets, and more.

Every one of these centers around information technology and demands CIOs be ready to help create new business offerings. And guess what — performing the same old packaged app shuffle is a recipe for irrelevance. Simply put, the eye-wateringly expensive packaged app game was fine for a time when IT meant centralized execution of easily standardized processes, but today’s economic environment calls for a much less expensive way and much more responsive way to create IT-enabled business functions.

Feebly offering up the mantra of “business benefit” as a shield to avoid confronting technology trends won’t protect you from other parts of the business demanding new  IT-enabled market offerings. Refusing to countenance open source as part of your solution portfolio is like a marketing person, circa 1995, rejecting the new-fangled Internet, citing his or her need to focus on “business benefit” marketing programs like TV spots and direct postal mail. Judging open source as a kind of second-rate enterprise app and concluding it falls short of the standard misses the point.

In a world where IT is central to how businesses run, implementing packaged apps to standardize your functioning by definition consigns you to middle of the pack — because packaged apps are designed to cover one standard deviation from the center of the bell curve: that’s how software companies make money — delivering middle-of-the-road applications. Being as good as the rest of the pack was OK when IT meant support for the real business of the company, but is a death threat in today’s economy.

I found this discussion of the “long tail of enterprise applications” particularly germane to the future of IT, which is why I discussed the theory of long tails in my last posting. The posting argues that following the traditional IT rules and process makes IT organizations unable to recognize or respond to real end user IT needs. The specific analogy the poster uses is the rise of the PC, in which end users smuggled the puny, RTFM, unapproved machines into the enterprise in order to take advantage of their powerful utility. It’s no accident I used that example in my posting — it provides the most vivid evidence of what happens when unstoppable end user demand confronts unbending IT resistance. End users always win in this battle.

In a world where digitization and information processing is pervasive, information technology has to get less expensive. If you don’t consider evaluating open source as an option to drive down costs, you are neglecting your charter as head of technology. InfoWorld just released the results of a survey they did which showed that 36% of respondents using one or more open source products used it in mission-critical applications; for those respondents with 100 or more open soruce products in use, the number using open source in mission-critical applications rose to 61%. What do they know that you don’t?