Will central IT allow independence?

Dear Bob …

I have a topic that I’d like to see you cover some time … service models. There are 3 that I know of:

1. Operational excellence

2. Product leadership

3. Customer intimacy

Most corporate IT shops run the operational excellence model … deliver

desktop systems that are the same everywhere, run servers that are the same

everywhere. Minimizing overall cost (or at least “first incurred cost”) seems to

be a hallmark of this model.  

I work in an “embedded” computing group. We are not part of the

intergalactic IT group that supports desktops and servers and the like. We work

for R&D researchers, and as such have to know a lot about their business,

have to offer customized services and products that change quickly due to the

nature of R&D. I’d say that we are in the “customer intimacy” service model.

I write all of this because there’s a study underway by the global IT folks

to pull in all “embedded” computing. They claim that they know how to manage

different support models, but all we ever see is the “one size fits all” mantra

that the operational excellence service model is known for.

Since I have a deep interest in how this turns out, I’d like to hear your

opinion on service models, and especially any history/advice you have on

organizations trying to blend service models.

– Being intergalactacized

Dear Warped …

I’ll take up the service models question in a future Keep the Joint Running. The other question you asked – is it possible for a central management group to support multiple service models – is worth some time and energy on its own.

The theoretical answer is of course. It is possible, in that nothing about it violates the laws of physics. It does, however, violate the laws of organizational dynamics, which means it’s highly unlikely to happen in practice.

The reason is connected to how the person in charge defines success. Each service model is connected to a different definition – one reason among many that the many executives who tell their teams, “We’re going to do all three!” invariably tear their organizations apart at the seams.

When central IT says it will successfully manage to multiple service models, it’s saying it can accommodate different definitions of success in different parts of the company. There’s no reason why not, except that it’s so hard to explain as you go up the chain of command. “Well, yes, the IT group in Altoona does spend quite a bit more in desktop support than the one in Walla Walla, but that’s okay because Altoona focuses on customer intimacy while Walla Walla focuses on operational excellence.”

“Altoona can be as intimate with its customers as it wants, so long as it doesn’t spend more than anyone else,” is the likely response.

Clayton Christensen discusses this at some length in The Innovator’s Dilemma, and concluded that it’s the single most important reason great companies fail at innovation: They measure the success of the innovation using the same criteria as they measure their traditional business. The ones that succeed at innovating are the ones that take the team that’s supposed to innovate and spin it off, viewing it more as a funded venture than a reporting business unit.

Which gets to the answer: If central IT views itself as a “holding company,” it can manage multiple service models, but it will be tricky because it will have to figure out how to deal with each “business unit” on its own terms instead of on a single, shared set of criteria.

All in all, I wouldn’t bet on it happening that way.

– Bob