What happens when we jump from one task to another while working? That’s not multitasking exactly. It’s task-switching. And the general belief is that it goes like this: First, there’s an interruption. Sometimes the interruption comes from the environment, like a push notification for a new email message, and sometimes it’s self-driven, like the desire to take a break. Whatever the case, it takes us a moment to reorient as we switch to the new matter at hand. Let’s call that new matter task B. Eventually, we decide to end Task B and maybe go back to the first task, task A, although we might rendezvous a little in something else along the way, task C. Finally, we land back at the original task, and it takes yet another moment or two to reorient and get back into it.
How much time do you think that whole process takes?
Unfortunately, research findings on the topic don’t align. It could take seconds, or it could take almost half an hour!
If we look specifically at task-switching related to email, a type of interruption that’s probably very familiar to most knowledge workers, the frequency at which it interrupts us (or we let it interrupt us) is crazy. Even if the interruptions are short, they add up. Office workers reacted to the majority of their incoming emails within 6 seconds of it arriving (Jackson 2001). Then it took them on average 64 seconds to resume work. Checking email was estimated to cause 96 interruptions in a typical 8-hour work day, which adds up to 1.5 hours per day reorienting.
I need to mention that Jackson’s research is a little old. The paper was published in 2001. But ask yourself whether office workers receive more email now than they did then, and whether office workers have changed their habits regarding how they react to email since that time.
Another team of researchers (led by Gloria Mark, 2005) noticed that getting sidetracked by other tasks cost 25 minutes before people returned to their original task, and that they rendezvous’ed with an average of 2.3 other tasks before getting back to what they were doing.
In some research from 2007, two researchers (Iqbal and Horvitz) noticed people spent 10 minutes on task-switches caused by alerts, such as email notifications, and another 10 to 15 minutes doing other stuff before they got back to the original task. In their research, 27 percent of all task-switching ended up in more than 2 hours of time doing something else before people got back to their original jobs!
What happens when people are told not to task-switch? In an experiment (Gould 2013) in which subjects were asked to complete computer tasks and without jumping around, about 60 percent of people did it anyway. Each switch took them on average only 16 seconds, and the primary task only took around 32 minutes to complete. But most people couldn’t bear not to interrupt their own work and look at something else, even if it was only for 16 seconds. On average, they did it 12 times in that half hour or so. (The median number of times subjects switched away was 6, indicating there maybe have been a few high outliers). Multiply 16 seconds by 12 times, and you get more than 3 minutes of wasted time in 32 minutes. That’s 10 percent of the time that people were working!
In studies, there’s no strong consensus about how much time we lose when we jump from one task to another or interrupt our own work to take a momentary break from it. But it’s clear that people do it.
The matter of whether it’s wasteful or unproductive to task-switch is hard to say definitively one way or the other. Like most things, it’s probably the case that it’s good for us and our work in moderation.
A few studies prove that interruptions don’t actually make us complete tasks any slower (Speier 1999, Zijlstra 1999, Mark 2008). It seems to be the case that a few interruptions actually make people work faster, as they compensate for the lost time they know they are experiencing. It may also explain why many people work best on deadlines. There’s a little bit of pressure to get the work finished. But moderation is key. Gloria Mark has proposed that although the occasional interruption might give us a bit of a time crunch and light a fire under us to hustle, there is probably added stress and frustration, and that the work requires additional effort. In other words, it’s less efficient and might then require more time to recover from the work.
Barley, S., Myerson, D., & Grodel, S. (2011) E-mail as a source and symbol of stress. Organization Science 22 (4): 887–906. Retrieved Jan. 22, 2016 from http://people.bu.edu/grodal/Email.pdf
Gould, S. J. J., Cox, A. L., & Brumby, D. P. (2013). Frequency and duration of self-initiated task-switching in an online investigation of interrupted performance. Human computation and crowdsourcing: Works in progress and demonstration abstracts AAAI technical report CR-13-01 (pp. 22–23).
Iqbal, S. T., & Horvitz, E. (2007). Disruption and recovery of computing tasks. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems – CHI ’07. Retrieved February 20, 2016 from http://research.microsoft.com/en-us/um/people/horvitz/chi_2007_iqbal_horvitz.pdf
Jackson, T., R. Dawson, & Wilson, D. (2001). Case study: Evaluating the use of an electronic messaging system in business. Proceedings of the Conference on Empirical Assessment in Software Engineering. Retrieved February 7, 2016 from http://homepage.ntlworld.com/cotwj1/pdf/Ease%202002%20Jackson.pdf.
Mark, G., Gonzalez, V. M., & Harris, J. (2005). No task left behind? Proceedings of the SIGCHI Conference on Human Factors in Computing Systems – CHI ’05.
Mark, G, Gudith, D, & Klocke, U. (2008). The cost of interrupted work. Proceedings of the Twenty-sixth Annual CHI Conference on Human Factors in Computing Systems – CHI ’08. Retrieved February 19, 2016 from https://ics.uci.edu/~gmark/chi08-mark.pdf.
Speier, C., Valacich, J. S., & Vessey, I. (1999). The influence of task interruption on individual decision making: An information overload perspective. Decision Sciences, 30, 337–360.
Zijlstra, F. R., Roe, R. A., Leonora, A. B., & Krediet, I. (1999). Temporal factors in mental work: Effects of interrupted activities. Journal of Occupational and Organizational Psychology, 72, 163–185.
Image by Acflynn, CC.
9 thoughts on “How Much Time Do We Lose Task-Switching?”
[…] (or task switching) cause you to lose anywhere between 15 – 25 min per interruption. If you are a SAHM or side hustler working from home with the little ones around … well […]
[…] people are in the meeting, but that’s only a portion of the time spent. Research shows that switching one task to another costs between 10 minutes to 2 hours. Meetings may interrupt productivity, but despite the cost, the data still shows a natural tendency […]
[…] from one task to another is the silent productivity killer. A report claimed that on average employees lost around 1.5 hours a day from its 8 working hours on […]
[…] Other investigations into interruptions and task-switching have found that on average, it takes about 25 minutes for a person to get back to their original task once they switch to another. And once interrupted by another task, the average worker ended up doing 2.5 other tasks before getting back to what they were originally doing. It takes more time to do the same task, and more unfinished tasks are added to pile. […]
[…] you constantly switch from task to task, you lose your focus and are less productive? That's called context switching or task switching and it's a massive productivity […]
[…] the extent of the problem. They found workers tend to switch between tasks on average every three minutes due to excessive interruptions. What that amounts to is a lot of lost productivity and […]
[…] are you getting?”, the answer was 10 (a day, a week???). While this may not seem like a lot, various research studies have discovered that task switching can cost up to 25 mins per task switch. This company was a […]
[…] you know that every time you swap between tasks, you lose time? Switching back and forth between tasks loses around 20 minutes while you need to refocus again. […]
[…] research showing that office workers reacted to most of their incoming emails within 6 seconds. Speedy! Work resumed over a minute (64 seconds) later, with emails found to interrupt work 96 […]