Lots of people are starting to question the degree to which people can, or should, multitask. For example, Scott Berkun’s recent Ignite Seattle talk was a version of an essay on the price we pay for our increasingly multitasked lifestyle. In that essay he writes:
It’s true that the hunt and intensity of multitasking can be fun — there are thrills in chasing things, physical or virtual, but most evidence shows we perform worse at all things multitasked. Despite how it feels, it appears our minds don’t work best when split this way.
Agreed. I’m lucky enough to be able to block out a lot of distractions and interruptions, and to spend an unusually large fraction of my working life in a state of flow. To the extent that I’m able to get a decent amount of quality work done, I tend to cite long periods of focused concentration as the reason why.
Even so, I’m in the habit of periodically questioning all my habits, including that one. So here’s a contrary view to consider. When I interviewed Mary Czerwinski on the subject of multitasking, interruptions, and context reacquistion, she made a fascinating observation about her teenage daughter, and tied it to some research findings. The observation was that her daughter has learned to operate in group mode, and that the groups she belongs to optimize themselves by moving tasks around to the members with the right knowledge, skill, and inclination for each task. The research finding was that although the resulting multiasking effect is suboptimal for the individuals and clearly damages their productivity, it can be the optimal way for the group to achieve its goal.
Since we don’t really have a choice about whether to multitask or not, the real issue becomes: What’s the right way to do it? The answer may be very different depending on whether you’re optimizing for individual or group productivity.
15 thoughts on “Multitasking tradeoffs: individual versus group productivity”
We have seen exactly this type of productivity effect on my team. We have worked together solidly for the last 8 years, but on and off before that for the last 18. There are four of us that are technical with one mainly administrative/business, and we all know each other’s strengths and weaknesses/likes and dislikes. Its amazing to see the productivity of the team breaking down tasks, routing them around and executing. Individually we block sometimes waiting on another to finish something up, but that break is also necessary to maintain sanity when the collective goals we are working toward are complex. The other thing we often do is individually branch out of the taskings we are working on to explore and then bring the results back into the team. That way we don’t stagnate under heavily focused implementation. Although that time is “lost productivity” against the goals immediately on our plate, that freshening effect is like pumping pure oxygen into a room…the learning brings with it inspirations which produce more energy on the tasks we have to do. I believe the net effect is maintaining high productivity without burning out.
I like the distinction: planning for group productivity or individual productivity. I do have a question though: when we are in group mode, we give tasks to the person with the right skills, but when they are working on the task, are they in a state of flow? At some point of time, being in a state of flow is essential do a good job of at least certain types of tasks. Tasks like writing, programming, etc are best done in a state of flow. Perhaps admin tasks (I am just guessing), or other tasks that have coordination (between people, agencies) as the main work goal, may be done efficiently without getting into a state of flow.
The two comments together enrich the basic picture. You’d want to optimize the group at the expense of individuals when:
(a) there is diversity in the group’s capabilities so that some folks do a given task much better than others, and there are consequently big gains from matching tasks to people well, and
(b) when the interruption costs to individuals are not too high, especially to key individuals who may be bottlenecks on the group’s performance.
All of this is still in a static view in which the definition of the tasks and therefore the effectiveness of members in undertaking them doesn’t change as a result of the conversation. But sometimes interruptions change your idea of what needs to be done. That brings in a whole additional level of tradeoffs in which dynamic task redefinition effects of constant updating can be either net gains or net losses.
In the end, I think the best advice on questions like this is not what is the best general policy, but rather what are the right questions to ask in specific work contexts.
Economists have a theory of Comparative Advantage to explain how it’s always beneficial to trade between countries, simply because they’re different.
I think the same theory applies to trading between people (except maybe for identical twins!) – that is, in specialising in different activities to achieve a common goal.
It’s always better to work together, it’s just a question of how much you do it – and Comparative Advantage gives us the maths.
A younger employee I worked with multi-tasked by working in 4 MS Word windows at once, cycling between docs. He claimed this was more productive than focusing on docs in single-file. I agree.
Checking email on my Blackberry while in a meeting is a way I personally multitask, and combined it’s clearly more productive (faster response, etc.)
When a goal isn’t clear, I find it good to multitask in order to gather more resources related to the outcome. Whether that’s talking to another person, a second monitor to help task switching or a tab load of Google searches. It’s when that gathering steps out of context that it becomes detrimental. Real-time communication (IM, Phone, Group) situations make it easier to be placed out of context than say email or a letter, why I believe it’s important to manage your attention and communication systems (not to mention teams) in order to stay goal orientated.
Individual and group productivity may well be best served with something like attention modes that allow you to context switch where your attention is, limiting or encouraging interruptions to certain individuals or groups. If I’m say working on a project for a couple of hours, I could add members to that and set communication levels based upon that. Whenever I set myself into that project mode, only members of that may communicate with me with say IM while having email and other notifications turn off. Another attention mode being a general mode free-for-all for when I’m freely available. I vaguely remember hearing that the new windows communication server (or whatever it’s called) is supposed to do much of this for teams? Of course in person when I might be in innovation mode I might go and sit outside in the sun. :)
The hard part is deciding how to clarify your goals and whether researching or delegating to another is more productive. Is it better to say ask someone you think may have a resolution or to find and learn that yourself. Ultimately that comes down to knowing yourself and the complexity of the task at hand. Knowing your skills as well as those you associate with.
I’d have to weigh up how difficult I perceive a task, and based upon that and knowing my own limitations, delegate accordingly. Managing my attention modes based upon that delegation. The problem is keeping aware of the state your in. :) It can be easy to get lost in the moment. It all comes down to self-awareness and what we value most, as to where we choose to place our attention.
Re: #5 (Jon Williams) – I have to admit that I find it hard to believe in both cases that the multi-tasking described is _more_ efficient – though if you have results to prove it, all the power to you.
In the first case, unless the user already has ADHD/ADD or the documents are all inter-related(i.e. requirement documents in which traceability linkages are being created between a host of requirements across the docs) the issue of context-switching time would seem to out-weigh any benefits of “keeping it fresh”.
In the second case, can we (humans) really respond to emails and still be fully attentive and _engaged_ (i.e. responding, not just listening) in meetings. If the emails were all one-liner, yes/no questions – perhaps this could be done between people speaking in the room. But otherwise, from my own experience I’d suggest that perhaps more efficient would be to not even be IN the meetings in the first place. Simply request the minutes later or ask someone about particular points in which you were interested. Not only will you be able to do more than email during that time, but you won’t risk giving the other meeting attendees a sense of distance or disengagement on your part.
Finally, personally I’ve found the common suggestion to simply for-go reading email in the morning before finishing one or two critical tasks on my mind pays dividends every time in productivity and focus.
This is just a personal observation, but I have not really found multi tasking to be of much help.
This may be a very simplified example, but I think it still covers a large range of related activities. Let’s say I am writing an email which happens to be an answer to something simple, I might decide to multitask by istening to a podcast at the same time. I might now complete the email and also listen to part of the podcast in the same time, thinking that I have gained productivity. But my attention was not totally focussed on the email (neither the podcast), so as the email conversation goes on (when I get a reply, and I have to reply back), I might actually have to reread what I had written earlier, resulting in greater cycle time. However, had I focussed just on writing that email, I would have retained what I wrote and would not have had to spend time rereading my original mail the next time.
Again, this is just my personal observation, which has led me to believe that productivity gained by multi-tasking evens out and in the long run there might actually be a net loss.
Or perhaps some people are better are multi-tasking than others.
Showing that multi-tasking isn’t all that efficient and Why multitasking sucks?