A plan for centralizing the service desk

Dear Bob …

I’m the corporate CIO of a semi-big company. We’re organized into semi-autonomous lines of business, each of which started life as an acquisition. Each has its own application development group that reports within the line of business, with a dotted line to me. I’ve gradually been centralizing the parts of the IT service catalog that seem logical, like e-mail provisioning and administration.

[ Also on InfoWorld, Bob offers some suggestions for setting IT support on the right track in “When the help desk malfunctions, metrics are a reliable root cause” | Get sage IT career advice from Bob Lewis’ Advice Line newsletter. ]

Up until now the centralization has been successful — we’ve saved money, and while there’s been some grumbling, it hasn’t been severe.

The service desk is next on my centralization list, and suddenly everyone is up in arms. I’ve been caught completely off-guard. This seemed like a no-brainer to me. I figured it would be like consolidating parallel call centers, which lets companies deliver similar service levels with a smaller combined staff because the law of large numbers smooths out unpredictable spikes in demand.

What did I miss? And what can I do about it now that I’ve already announced my intentions?

– Wrong side of a hit fan

Dear Splattered …

What you missed was everything important about service desks other than cost and the two official metrics: the time to initial response and time to resolution.

It’s one of the hazards of measurement — anything you don’t measure you don’t get, no matter how important it is.

The service desk is where the relationship between IT and the rest of the business happens, far more than any other type of interaction.

You deal with executives and managers when the subject is strategy, priorities, and whatever else you talk about. Project teams interact with project stakeholders. The service desk works with end-users when they need help right now.

End-users don’t just interact with the service desk on a process basis. They interact with human beings, who by now know them as individuals, understand their idiosyncracies, and have a pretty good handle on how they use technology day in and day out.

Service is personal. It’s tailored to each individual. End-users don’t have to take a lot of time explaining things.

You’re going to take all of that away, turning the service desk into a set of impersonal and less-convenient processes. And if, as seems likely, the lines of business are geographically dispersed, you’re probably going to lay off a lot of those individuals with whom the end-user community have bonded.

Of course this will generate resistance.

What can you do now? I suggest you start by listening, and not only to your peers. Get a handle on the concerns of everyday, ordinary end-users, power users, supervisors, and even middle managers. Acknowledge everyone’s concerns and let them know you’re going to modify your plans based on what you heard.

Specifically, what you’re going to do instead is virtualized centralization. You’ll centralize support for some of the most common and easily handled incidents, such as password resets, but leave enough service desk technicians in place locally that the customized service everyone values so much won’t be lost in the shuffle.

Or apologize for the inevitable disruption, doing what you can to persuade the objectors that the company’s need to cut costs has to override the disadvantages, which you don’t deny and will do your best to mitigate.

Demonstrating the ability to learn, and to recognize a mistake, doesn’t signal weakness. It signals confidence.

– Bob

This story, “A plan for centralizing the service desk,” was originally published at InfoWorld.com.