Should You Reserve Some IT Capacity?
It's always a bad idea to plan more projects than your IT department has the capacity to carry out. But should you plan substantially fewer, keeping some IT work hours in reserve for contingencies?
That's what Gartner analyst Robert Handler advises. "In theory, if everyone came to the table during budget time with information on all the systems they need, it might be possible" to plan to work at full capacity, he says. "But a week or two after budgets are done, there are already a lot of requests for new stuff. We're in a complex world and there are changes constantly coming from markets and legislature. They destroy the predictability of projects."
Few IT operations are effective at dealing with the unpredictable nature of their work, says Handler. So he looked at other fields for inspiration. He found it in new product development. "Their response is to maintain reserve capacity for uncertainty," he says. He believes IT departments should do the same.
"Some business leader will say, 'We need this project to do business,' and if the CIO says, 'No, we can't, we're at capacity,' the answer will be, 'Then we'll get it elsewhere because we need it!' I suggest you reserve capacity for that situation." IT's goal should be to run at 80% of capacity, reserving the extra 20% for "things that come out of nowhere," he says.
Todd S. Coombes, executive vice president and CIO at ITT Educational Services, disagrees. "I've worked in environments where you set up contingencies, and I prefer to work based on historic data," he says. "If my data from past projects tells me I need to reserve a certain amount of time for unplanned activities, we'll work toward that, rather than assume we need to build in an extra 20% thinking things might go wrong." When people know they have that leeway, they tend to use it, he explains, adding, "I like to have things a little tighter."
Coombes says he uses detailed planning of every IT employee's time, and then has them track their activities as projects progress. "We can see historically what they actually spent their time on," he says. "It's kind of a feedback loop of planning and setting our capacity target, collecting actual information and then studying the data." That process allows for increasingly accurate planning.
This way, Coombes can plan for the unexpected on those projects that warrant it. "I don't know exactly what unexpected thing will happen, but historically I know it's going to be something," he says. "So we will build that into our capacity model. But it's based on what we know to expect." Minda Zetlin
Sign up for Computerworld eNewsletters.