Think Multitasking = Greater Productivity? Think Again
A couple of posts back I asked whether working long hours was good for productivity. (Spoiler: It isn’t.)
Here’s another thing that isn’t – multitasking.
All it takes is a common sense calculation:
I’ve worked with at least 500 companies and organizations since I went into business in 1992. I can’t think of one of them that doesn’t engage in multitasking i.e. where a pool of people is spread across and works on a series of projects / activities (‘things’).
Generally then, every person multitasks i.e. spends time on more than one thing. And I’m not talking about just two or three things. It’s much more likely to be 7-10 things. I’ve seen people multitasking across twenty things.
But multitasking is catastrophic for productivity as the following common sense example shows.
Let’s say I’m running a project. The project consists of a number of jobs that have to be done. Let’s say one of those jobs is estimated to be 10 person-days (PD) work. Charlie’s assigned to do it and Charlie is not multitasking – he’s available full-time, 5 days per week (dpw). Then the job will take two weeks.
Now suppose Charlie is multitasking and he’s only available 1 dpw. Then the job will take ten weeks.
But there’s more.
There is now the extra time involved in Charlie putting the 10 PD job down, picking it up again a week later and getting his head back around it again, back to where it was when he put it down. (There isn’t a lot of research in this area but google on ‘the cost of task switching’ for some representative figures.) Based on this research, an estimate of 15 minutes to get his head back to where it was wouldn’t be unreasonable. So that’s an additional 10 [weeks] x 15 minutes that has to be added on to the initial 10 PD.
And that’s not all. There’s another catch. The 1 dpw that Charlie is giving to our project is – most likely – not going to be one full day, as in, for example, Monday. It’s much more likely to be something like this: a couple of hours on Monday, half a day on Tuesday, nothing at all on Wednesday, an hour for a meeting on Thursday and then a bit of a flurry on Friday to hit some mini-deadline or milestone. So instead of the initial 10 put-down-pick-ups, we’re much more likely to have 4-5 put-down-pick-ups per week. Over 10 weeks, that’s 40-50 put-down-pick-ups. At 15 minutes per put-down-pick-up, that’s 10 – 12½ person-hours added on to the initial 10 PD – so this is very non-trivial.
But let’s pretend this doesn’t happen – if only because we’re not quite sure how to measure it – though clearly it does. The difference between Charlie working 5 dpw (not multitasking) and Charlie working 1 dpw (multitasking) doesn’t seem that serious. On a big project you might not even notice it. Hey, on a small project, with Charlie sitting beside you or just over the partition, you’d see him on his computer, making phone calls, going to meetings, you might think ‘Charlie’s doing my stuff’. Yet, this one small thing – the difference between 5 dpw (not multitasking) and 1 dpw (multitasking) – can potentially cause an 8 week delay on this project [the difference between 2 weeks and 10 weeks].
And remember this is just one job in your project. Is this happening on all jobs that Charlie is doing on your project? Of course it is. And is this happening on other jobs on your project? Most likely it is, since most everybody is multitasking.
The resulting stretching / delay on your project can be colossal.
Or to put that another way, the amount of overtime the team will have to work to stop the stretching from happening is colossal.
So why does everybody do it?
Is it because nobody has thought about the problem? Or people think there’s no alternative? Or what?
Because of course there is an alternative. And it’s this:
1. Prioritize the list of projects that have to be done.
2. Flood the most important project with people who are not multitasking (or who are doing as little multitasking as humanly possible) and aim to get the project done as quickly as possible.
3. Working down the list, do this for the following projects until you run out of people.
4. Don’t start the remaining projects until people become available from projects as they complete. The remaining projects will still be done quicker than if you had been multitasking.
5. Do this and you’ll be amazed at the quantum leap in productivity you’ll achieve.
---o0o---