Thursday, October 25, 2012

Enterprise Computing’s Costly Habit of Long-Term Planning

Enterprise computing is still clinging to long-term planning even though it no longer makes any sense in today’s business world.

I know the long-term planning vibe is still very much alive from a recent Secret CIO column in Information Week. John McGreavy devotes an entire column to a rant against Google, eventually concluding:

Google, you have some great products. . . . but it will take some adjustment in thinking and approach to conquer the enterprise. An attitude adjustment wouldn’t be a bad start.

Poring over the fine print, the supposed attitude problem McGreavy finds in Google is its reluctance to make specific product promises more than six months out. This is what has McGreavy so appalled. If you can’t make commitments five years in advance, he goes on to suggest, you don’t belong in the corporate world.

People who aren’t privy to the world of corporate IT must have a hard time imagining the sense of entitlement that this view represents. What kind of delusion could lead anyone to assert the privilege of separating their efforts from reality for five years at a shot? It must be something like this: “I am the king of the world. If it takes five years for me to get out of bed and come down to breakfast, the world will just have to wait.” This suspicion is confirmed when you look at what corporate IT has done in the last five years and compare it to what Google has done. Just one point of comparison: The IT department is still forcing workers to use Microsoft Windows XP, an operating system that hasn’t been officially supported for nearly five years. Google, in the same period of time, has coded a new operating system from scratch and has more than 100 million people using it. So who has the more realistic view of the world of work? Indeed, the long-term plan is a big part of what prevents corporate IT from getting much done.

Of course, it is not just Information Week. Hewlett-Packard had people chuckling this month as it laid out its five-year turnaround plan. It was so funny because in the computer business, “five years” serves as a euphemism for “never.” The executives at Hewlett-Packard might as well have said, “HP will be profitable again when hell freezes over.” Despite the awkward presentation, Hewlett-Packard is serious about its five-year plan — for now, anyway. No one seriously expects the plan to survive intact for more than a few months.

You can’t plan your work five years in advance if you’re working in a rapidly advancing area of technology. Before you finish writing a plan, the world changes enough to render the plan obsolete. A five-year plan turns into a five-week plan. In practice, the long-term plans in computing are revised continuously, at enormous expense. So where did the five-year plan in information technology come from?

It makes more sense if you look back to the 1970s and the behemoth computers that dominated business computing at that time. A computer system might cost $20 million in today’s money to purchase and more than a year to set up and test. It would be expected to last at least 10 years. You can’t make that kind of investment without a long-term plan.

But a comparable computer of today costs about $5,000. You can easily sell it if you can’t find another use for it after you’re done with it, so you are not making a major commitment by buying it. The manufacturer can deliver it overnight if you are in a hurry. It can be set up and tested in less than a day. There is no need for long-term planning, no benefit in it, when technology comes in such small increments. Long-term planning in corporate information technology is a habit that has outlived its usefulness, yet it carries on, and it forces the entire enterprise to slow down to keep pace with it.

This is more than just an academic concern. If the multi-year information technology plan has become the anchor that prevents the enterprise from getting out of the harbor, then economically, one of two things will happen: the enterprise will cut the anchor loose, or it will sink where it sits. When the storm comes, my feeling is that most corporations will decide they want to stay afloat.