Multi-tasking and efficiency

Tom Breur

21 December 2013

Recently, me and my esteemed colleague Michael Mahlberg ran a workshop where we played Henrik Kniberg’s “name game.” Always good fun, and lots to learn. It never ceases to surprise me how different people can learn so many different things form the same simulation they are participating in.

The objective of the name game (I assume) is to help people experience the pernicious effects of multi-tasking. We also like to use the game so that people can experience the power of measuring, without making it overly obvious.

When participants physically record the start and end times of their task, interestingly, they don’t immediately “see” what is happening. First they take turns, finishing their work piecemeal. Then, in the second round, they take turns but finish the entire task. ‘Naturally’ this goes a lot quicker. However, the fact that in the latter case everybody –except the first person– starts later, yet (almost) all finish earlier, is often not obvious while writing down the start and end times of the tasks.

Our common-sense notion is that if you start earlier, you should also finish earlier. Not so in the name game. During the first part of the game, because people take turns adding another letter, they are actually multi-tasking, and the task switching causes significant delays. 

Participants write down the start and end times of their task. Only when they look at the data, does it become ‘obvious’ that although people started later in the second part of the simulation, their “work” gets finished so much quicker. The notion “Let’s already start this project, so we can work more efficiently” is often based on a false premise that utilizing idle time will make work go faster. In contrast, it tends to create more Work-in-Progress, requiring more task switching. Multi-tasking seems to be a somewhat hidden time sink…