Own a system for all time your Cash Loan Company Cash Loan Company faxless hour online website.

Clarke Says, Multi-Tasking Is Evil, I Agree

by Hal on May 15, 2008

in Theory of Constraints, agile, lean

If you're new here, you may want to subscribe to my RSS feed. Thanks for visiting!

I won't bore you with all the references to how multi-tasking produces waste. But do understand, the company policy to have very high utilization of staff creates the requirement for multi-tasking. Full utilization is not sustainable. Until you can lower utilization, thereby creating slack, you won't be learning and innovating. You can't be lean.

Clarke Ching, writing for Sticky Minds, uses a simple exercise to show just how evil multi-tasking is. Do the exercise for yourself and then have your boss do it. It goes like this:

Make three columns on a page for the three projects X, Y and Z. For the first project (column) you'll write the numbers 1 – 10; in the second column write the letters A – J; and in the last column the Roman numerals I – X. You'll perform these three projects twice. Have someone time you.

Project X Project Y Project Z
1 A I
2 B II
. . .
. . .
. . .
10 J X

The first time performing the projects do it in multi-tasking mode by taking the first step of each project by recording the first character for all three sequences, then the second character for all three and through to the tenth characters — row by row — until all three projects are completed.

On the second pass complete one project before going on to the next by recording one full sequence in a column, all the numbers, then all the letters, finally all the Roman numerals — column by column.

Now, compare your times. Also, note if you made any errors along the way.

I've used a number of exercises and examples to show the mal-effects of multi-tasking, however none are as powerful AND as easy to demonstrate as Clarke's three-column exercise. I'll post my results in a comment to this posting. Please share the results and reactions that you got. Thanks Clarke!

{ 20 comments… read them below or add one }

1 Hal May 15, 2008 at 10:12 pm

I did the exercise in multi-tasking mode in 52 seconds. I made 1 mistake. In the second round doing one project at a time I did it in 18 seconds without any mistakes. While I’ve known that multi-tasking is less productive, I wouldn’t have said that it is 3 times less effective before doing this exercise.

Notice, when not multi-tasking one project finished on average every 6 seconds. Compare that with the record in multi-tasking mode where the first project finished at around 48-50 seconds. Also note, that when multi-tasking there is much more work in process.

Some people have said that multi-tasking makes you stupid. No…but, it certainly requires much more attention.

2 clarke ching May 16, 2008 at 2:26 am

Glad you liked it Hal. I first heard this example on the Critical Chain yahoogroups list but I can’t find the original message. It does work doesn’t it!

3 Claude Emond May 16, 2008 at 5:21 am

Fantastic exercise Hal. I’ll use in my workshops.

Very, very effective demonstration. Simple things make for better understanding hehe

4 Diana Hutchinson May 16, 2008 at 8:52 am

My speed difference wasn’t so much as Hal’s — 26 sec vs 34 sec. But I could definitely tell the difference, and it did cut off about 1/3 of the time not to multitask.
I’ve done a similar exercise with colored poker chips. One time we had to sort one red, one blue, one red, one blue. The next time it was all reds then all blues. That was also a significant difference.
Thanks for sharing this simple exercise, I think I’ll get a few minutes to do it at our next PM meeting.

5 Glen B Alleman May 16, 2008 at 9:29 am

Clarke and Hal,
What is the suggested solution to this issue. One that can be applied here in the field. One that still maintains the financial goals of the project and the business stakeholder of labor absorption, client billing, or just plain labor utilization?

6 Hal May 16, 2008 at 9:54 am

Glen, in the AE space we notice that people pay so much attention to staying billed at least 100% of the time that they resist spending any time in conversations and work that could make the way they did their work more valuable to the client and the AE firm. I’m working with a top 5 AE firm at this time. We have encouraged them to use a PDCA approach to doing design work. By getting more people involved in the planning of the work and then following that up with “check” and “adjust” they are finding immediate learning and innovation that they hadn’t been getting.

Part of the PDCA approach they’ve adopted includes small batch design. This has led directly to less multi-tasking. PDCA in combination with small batches is getting their projects done faster with less resources and far fewer loop backs.

7 Hal May 16, 2008 at 8:00 pm

Diana, I’m much older than you. Even with practice I don’t think I’ll be a competent multi-tasker. But even if there’s just a 20% opportunity from stopping multi-tasking it’s worth pursuing.

8 Joe Ely May 17, 2008 at 3:51 pm

Great exercise, Hal…thanks to Clarke for the tool!

9 Glen B. Alleman May 19, 2008 at 9:16 am

In the space and defense domain, 100% billing is not a specific issue, but 100% flow of work is. Since the planning process assumes fungible resources at the work package level (these are project with 100’s to 500 engineers spread around the country), the problem of balancing the resources is left to the Control Account Manager and flows down to the Work Package Lead. Design is the primary work product for the first half of the program, so sequencing the work is not really possible in the same way as manufacturing, test, and flight operations.
PDCA is an option, but self organizing the WP Team is the most common, which due to the nature of the work turns into heavy multitasks. The phrase used to describe the situation is OBE (Over Come by Events). Once we get behind the planned hours absorbtion – due to normal overload – multitasking is the only way out.
Fine grained iterations solve some problems, but it’s a fact of life the multi-tasks is part of the space and defense execution process.
While Clarke’s exercise is cleaver, I’ve yet to hear an actual solution that can be put in place with measurable benefical outcomes. It’s restating the obvious that multi-tasking is less efficient, but still waiting for a better approach that can be put to work on “real” projects – other than the fine grained deliverables. At Rocky Flats, this was called Plan of the Day and supported by the Plan of the Week from the previouos week.

10 Don Miller May 19, 2008 at 10:17 am

At a nuclear facility to reduce maintenance backlogs, we developed teams to work mostly back shifts to reduce the backlog. we utilized one electrican, one welder,one mechanic and an I&C tech. they were self directed and multitasked so that only one of each trade was on the team.
this process worked very well for us. errors were minimal and backlog was reduced. If we did not multi task we would have needed at least twice as many complicating the planning process.

11 Hal May 21, 2008 at 12:30 am

Hi Don,

Nice hearing from you. Has it really been 7 years since we worked on the nuclear power station?

The idea of multi-tasking has more to do with one person who has 2 or more tasks going at the same time than 2 or more people doing one task. I’ve also seen great value in creating composite crews. I hear people say that work rules get in the way of that. It’s not my experience. Last winter I worked in Canada on a power station where we created composite crews. We too had great success.


12 Gil Friend June 26, 2008 at 1:10 am

Thought provoking and just what I needed. Thanks.

And a question:
You say “Full utilization is not sustainable.” Yet utilization is typically a key metric (and a key profitability driver) in a consulting firm. So how do you suggest thinking about and managing utilization? (Do you set utilization targets? And if so, at what level?)

13 Hal June 26, 2008 at 2:14 pm

Hi Gil,

Great question. Utilization is an indirect leading indicator of short term firm profitability. However, there are three situations that require slack capacity: ability to respond to opportunities, working on the business, and learning. Some firms reserve capacity for this setting lower utilization targets for staff.

One thing you can do is to adopt a make-ready approach for your work along with the rule not to start a task that is not in a condition to be finished. The combination will result in far less multi-tasking.

14 opleiding projectmanagement July 23, 2008 at 3:30 pm

It is actually what the zen masters are saying since centuries.

15 Alex September 17, 2008 at 11:06 am

very interesting post. that exercise is a really effective and simple way to prove the point about multi-tasking. and that is the most intelligent way to present an argument! you have great PM blog here. I also have a blog on project management, check it out if you get a chance http://www.santexq.com/blog. keep up the good work.

16 AlexJB September 26, 2008 at 5:57 pm

An interesting observation, and great ‘back of the envelope’ style demonstration. but misleading, and rather disingenuous in its simplicity.

There are a number of problems with the underlying premise.

1. premise: each ‘project’ could be taken from beginning to end without interruption if focused on exclusively. i think we can all agree that this is not true on real project work 80% of the time.

2. premise: zero flaws is the preferred outcome. this is not always true either – speed is only one of the sides of the iron triangle.

3. premise: each project is only done when it’s done. sometimes, there are sub-project milestones which, if achieved in a timely way, provide substantive value. you could argue that this is just a matter of re-sizing the “project”, so i’ll throw in…

4. premise: there is no value to a partially started project. in my experience, there are frequently times when having a project ’started’ has real and tangible value to the stakeholders.

as fun as it is to malign multi-tasking, i think that it’s unrealistic and pointless. i multi-task constantly, all day long, even more when i’m not at my job than when i am. while waiting for Quicken to boot up, i pet the cat or get a glass of water. or i’ll alt-tab to my web browser and trigger my RSS feeds to load so that i can flip over and read one at a time while waiting for some other operation to run. when a TV show goes to commerical, i go pee or pick up the conversation that i was having with my friend.

it’s not evil, it’s as natural as breathing.

17 Hal September 27, 2008 at 8:17 pm

Disingenuous? C’mon! Yes, we multi-task when making coffee and letting the cat out. We multi-task when listening to the news and reading our email. Mothers feed their babies while stirring the soup. Of course it’s like breathing. But, we’re not talking about that. We’re talking about professional work. Work that takes attention…often our full attention.

There are two big issues. First, as the exercise clearly shows, some of the work that we do will take longer when multi-tasking. Second, for those of us who do professional work — structural engineering, architecture, software programming — multi-tasking introduces errors that can have consequential results.

These are the reasons Clarke calls multi-tasking evil…and I agree.

18 ds April 3, 2009 at 6:48 pm

I have to say, I did it in both modes in *exactly* 39 seconds in each. I often have to multi-task in my job, so that may have something to do with my results. When I don’t have to multi-task, I finish tasks front to back and then move on, but when I reach a point where I need to wait for results to continue, I switch to the next thing, usually small tasks that require less focus & match the amount of time I will need to wait.

19 Hal March 15, 2010 at 4:07 pm

I just repeated the exercise with 8 people. The average first result (working on three projects at once) was 46 seconds. The second result averaged 21 seconds. The most interesting thing was that all the errors occurred in the task-switching round.

I just got back from working a week in Ireland. I rented a car. It had a stick shift. I drove for the first time in 20 years on the left side of the road shifting gears with my left hand reading road signs with the names of three up-coming exits alternating in Gaelic and English. It’s not quite the same as task-switching, but boy did one thing get in the way of the others. It took me 2 days to get used to driving on the left and shifting with my left hand. I never did get used to reading the dual language signs. I kept missing my exits.

20 Jeff April 15, 2011 at 12:20 pm

I’m working with a team of architects and engineers on a very large factory retrofit design. The project is composed of many dozens of small projects. We’re trying to work in small batches by planning in short cycles. One challenge is the contracted definition of done at milestone reviews. The team is conditioned to make the grade – literally to score well on the ‘report card’ they get at each review. They have, in effect, promised to work on everything all along the way and to finish nothing until everything is done. We can see how this is a problem, yet have been *unable* to negotiate a different approach with the customer. Yet we must.

Leave a Comment

You can use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Note: This post is over 5 years old. You may want to check later in this blog to see if there is new information relevant to your comment.

Spam Protection by WP-SpamFree

Previous post: Vroom and the “Capability Principle”: from sharing the project vision to successfully delivering projects

Next post: Read Robert Sutton’s Blog