I don’t know if you’ve noticed but the IT industry is gripped in a mania for centralization right now. If this were fashion, we’d be talking about the new black, but since it’s technology, I’ll have to settle for saying centralization is the new outsourcing. The big “C” is giving outsourcing some serious competition as the option of first resort when it comes to managing IT cost and quality issues.
Though it might not fly in the states of Kansas or Ohio, this is all part of the natural organizational evolution for IT. In the dim pre-history of computing, the only way to bear the substantial costs and demanding ministrations of the big iron was to have a centralized glass house. There were significant challenges in flexibility, effectiveness, and satisfaction, but it was the only game in town.
The next thirty years saw the big iron morph into departmental computing, desktop computing and the omnipresent network, capped off with that noisy and transitory fireworks display known as the Internet Bubble. As always, today’s solution is tomorrow’s problem. The flood tide of localized computing left the rotting fish stink of runaway costs and decaying reliability in executive suites across the land.
Wholesale outsourcing was the first instinct in many of those suites, but it delivered questionable value benefits and outright reliability failures as the complexity of external relationships and technology overwhelmed management capabilities and user tolerance. Into that void steps centralization.
Now a funny thing is happening on the way to centralization. Regardless of what level of the organization is talking, I almost always hear exactly the same plan: “Centralize everything below me, but what I do is too unique and too closely tied to the needs of my users to be centralized to any part of the organization above. Those clowns up there don’t understand or care about the realities of how we do business.”
CIOs say it about outsourcers. Business unit execs say it about corporate IT. Departmental managers say it about the business unit technologists, and the guy that just got hired onto the help desk with his degree in American Literature says it about the help-desk manager. It’s kind of eerie hearing the same words come out of so many mouths.
The only problem is that a dictum of “Above me localized, below me centralized” just doesn’t work as an enterprise organizational architecture for computing. Here I have to grit my teeth and admit that Nicholas Carr’s infamous article “IT Doesn’t Matter” actually contains a germ of a useful idea. You remember his basic tenet: IT has become so ubiquitous that it’s just a utility that offers no differentiation or competitive advantage.
Mostly obscured in that wild overreaching is the idea that some technology is becoming like a utility. Not all technology provides instant business benefit in a poof of cash and acronyms. Despite the current obsession with de-regulation, the most reliable approach to acceptable quality at acceptable prices for utilities is the regulated monopoly, a.k.a. centralization. The real trick for IT is to distinguish between technology that is sufficiently evolved to be considered a utility and the stuff that’s still emerging and requires hands-on, day-to-day, by-the-business management.
The architectural dictum becomes “Localized custom, centralized utility.” Cool mantra, but without some guiding criteria it only shifts the debate to definitions of custom and utility. Bill Clinton would thrive in that linguistic rodeo.
The localized custom approach has three leading indicators. The first is that your local use of a particular bit of technology varies from common usage either in volume, frequency, or criticality. Really big consumers of electricity have their own backup generation capabilities, for example. The second is accountability for failure. Reasonable people won’t blame you for a power failure (that’s the utility’s fault), but will blame you for a design flaw in your product or service (that’s your fault). The third is rarity of supply. If your organizational marketplace offers up no other feasible alternative for supply, you’re probably best served by the localized custom approach.
The centralized utility approach also has some leading indicators. The first is that the technology is well understood. We’ve got over a hundred years of telephone network experience. VoIP may change which external network we use, but nobody is suggesting we bring that external network in house. The second indicator is a number of equally capable technology alternatives. If they’re all equal, the overhead of variety is a tough sell. The third leading indicator is emerging industry or institutional standards or direction. If there’s city water and sewer at the curb, digging your own well and maintaining your own septic system may not be the best option.
Every organization is unique. All the prepackaged solutions in the world won’t seamlessly respond to this reality. We’ll have to roll up our sleeves and do the hard work of understanding our technology in our business contexts. With that foundation, a good set of ground rules and the right partnerships with the right vendors, we can find our unique organizational balance points between localized and centralized.
The opinions expressed herein or statements made in the above column are solely those of the author, & do not necessarily reflect the views of Wisconsin Technology Network, LLC. (WTN). WTN, LLC accepts no legal liability or responsibility for any claims made or opinions expressed herein.