Operating at Maximum Capacity is Actually Less Efficient

April 10, 2013

By Dan Markovitz

My friend Alex is an IT architect at a large telecom firm. His team is responsible for evaluating, testing, and supporting new technologies throughout the company. I probably don’t need to mention that they’re overwhelmed both by the volume of new tools they need to test, and by the unanticipated requests by the CIO, which have to be slotted into their daily work and their regular project flow.

At first, Alex’s team was viewed as a lightning-fast, responsive group. Now they’re seen as a bunch of molasses-like foot draggers. Same group, same skills, same commitment — only now they’re goats rather than heroes.

It’s not their fault. Blame it on management.

Managers hate idle time. When people aren’t working every minute of the day, managers see excess capacity, and they start a new project. Even if it’s clear that the new project won’t be completed immediately, they assume that starting sooner means the project will get done sooner — after all, if you can get a few tasks done today, you’re that much farther ahead.

But this assumption is false. Little’s Law, one of the classic formulas underpinning operations research and queuing theory, states that on average, cycle time is proportional to the size of the queue divided by the processing rate. If you keep cramming new items in the queue — customers at a coffee shop, candies on Lucy’s conveyer belt, cars on the 405 freeway at rush hour, projects in the IT architecture group — the time it take for an item to come out the other end takes longer. It takes a whole lot longer, in fact, because as you approach 100 percent resource utilization, wait times will make the line at the DMV look like an express lane.

The truth is, it doesn’t matter what department you work in — IT architecture, footwear product development, or customer service — if you’re juggling too many internal and external demands, the wait times lengthen exponentially. Notwithstanding managerial exhortations to multitask, human beings cannot do two cognitively demanding things at the same time. (Walking and chewing gum doesn’t count.) Each time you switch from one task to another, there’s a real and significant loss while your brain orients itself to the new task — and another loss when you go back to the previous task. These losses add up, until eventually people are griping that glaciers move fast compared to you. Moreover, remember that the slower a project progresses, the more likely it will be obsolete by the time it’s ready.

Here’s what you can do to avoid this problem:

  • Freeze 25 percent of your ongoing projects. (Which ones? A pareto analysis is a good place to start, as is a formal assessment of the business case for each project.)
  • Rigorously control the rate at which you start work. If you’re involved in a process that’s repeatable and somewhat predictable, match the incoming rate at which you start work to the rate at which you finish work.
  • Ensure that any new project is adequately staffed until it’s completed.
  • Don’t steal resources from another project to make headway on the new one.

It may take a while for your team — and your customers — to see the value in this approach. At first, they are likely to think that you are being less responsive. But once they see their projects moving to completion more quickly, they’ll become supportive.

Dan Markovitz is president of TimeBack Management, a consulting firm that radically improves team and individual performance. He is a faculty member of the Lean Enterprise Institute, and teaches at Ohio State University’s Fisher School of Business and Stanford University’s Continuing Studies Program. Markovitz formerly held management positions at Sierra Designs, Adidas and Asics.