Goldratt Research Labs / Multitasking Simulator review

Most job postings ask for multitasking skills as if it is a "good qualification". I consider it as juggling, how many balls can you flip in the air at the same time?

This has important repercussions especially in white-collar R&D works and project works. Most of us think that by putting projects into operation one after another and as early as possible, we ensure that they are delivered on time. However, white-collar personnel accumulates several projects at the same time, and they have to do something about each project every day for some reason or answer some questions, maybe fill out some reports. Moreover, some evaluate the personnel by measuring the arrival and completion time of these projects. Therefore, personnel are divided between projects, get stressed, experience quality or efficiency problems, and sometimes cannot meet the content fully or finish their task in the project late.

There are two main options in this simulator by  Goldratt Research Labs: computer or manual. 

It is recommended that we try with the computer first so that we will "learn the job", then we will know what to do when we try it manually or with our friends.

There are 3 projects that we can release to our team, and each of them can have 5 - 10 - 20 tasks. In one project, the tasks are letters of the alphabet, in another they are numbers, in another they are symbols. We can change the duration for each task, the margin of safety, the duration of distraction when switching between tasks, and the adaptation period when switching between projects. Thus, we can reflect on our own work environments, such as short or long-term assignments, similar or different tasks, and similar or different projects.

When you try with a computer, you can also change the error rate and the time it is required to fix an error for multitasking and single-tasking.

While the simulator is running, you can monitor it based on Gantt Chart or performance factors. 

If you start with 5 tasks, you get fast results, but if you start with 20 tasks, the multi-single difference becomes more obvious. 

After running for both cases, you can still compare the results as Gantt Chart or performance factors. 

You have 3 options when looking at the results: 

Planned multi vs single comparison,

realized multi vs single comparison, 

multi vs single with Gantt chart...

 

I have tried both computer and manual modes

Computer mode:  I set a 10% error margin for multitasking and a 5% error margin for single-tasking, without altering the current setup for times and safety margins for 3 projects of 20 tasks. 

3 projects X 20 tasks per project X seconds per task I actually planned to finish in 60 seconds.

Multitasking was completed in 80 seconds, 4 errors were made, and 0.75 tasks per second were completed. As a result, the completion of all 3 projects was delayed by 20 seconds (33%!). 

The distraction due to constant switching between projects costs 30 seconds, a loss longer than the total delay, hence the whole project is late.

When the computer performed single-tasking, it finished sooner than 80 seconds or even 60 seconds of the planned duration, it took just 47 seconds! Switching between projects only took 1 second. Thanks to focused work, only 1 mistake was made instead of 4 mistakes. The number of tasks completed per second increased from 0.75 to 1.28. We completed the projects early and will be able to start new projects in the remaining time!

At the planning stage, both methods looked the same... We were able to shorten the finishing time by 42% (from 80sec to 47sec) with a single-tasking mode, we reduced the internal open project stock by 67% (from 3 projects to 1 project), errors by 75% (from 4 to 1 error) ), we increased our task completion rate by 71% (from 0.75 to 1.28).

Manual mode:   I do it myself with the current setup for times and safety margins for 3 projects of 20 tasks. According to the plan, there should be no difference between the two methods, as we see in the computerized mode.

When I was in multitasking style, I was very careful, and I didn't make mistakes, but when I was in a single style, I relaxed feeling "safe", I made a mistake. 

My finish time is 40% shorter from 75 seconds to 45 seconds. Open project stock decreased from 3 projects to 1 project as expected. My task completion rate has increased by 67% from 0.80 to 1.34!

We also try it with my colleagues on the team. What is surprising is that some of them are more prone to multitasking. They made fewer mistakes and finished faster than the single style. They are known as hyperactive. 

I think the best approach would be to respect those who have similar working styles and form the right team by utilizing this simulator. Then let them choose the best fitting mode (multi vs single) for themselves. Please also note that multitasking is different than being multiskilled...