Employees who socialize between tasks, but not during, are more productive.
Even cubicle monkeys need a social life if they're going to do their job well, say US researchers.
Tightly knit groups of colleagues who communicate with one another frequently are more productive than those who are more isolated, the researchers have found. On the other hand, communication at the wrong time reduces productivity.
The researchers used electronic monitoring to tease apart the various types of interaction in the workplace and their differing effects. Such monitoring could improve how individuals and organizations work, but it raises issues about the extent to which companies monitor their employees' behaviour.
Many studies of communication within organizations, such as of who e-mails whom, have suggested that loose networks, in which people have few contacts in common, boost productivity. But these don't capture face-to-face, moment-to-moment communication, says Benjamin Waber of the MIT Media Lab in Cambridge, Massachusetts.
"People have formal structures and reporting relationships, but when you look at who's actually talking to each other you get a different picture," says Waber. "We can predict productivity far more accurately from these informal structures and behaviours."
Waber and his colleagues equipped a team of 23 employees at a Chicago IT company with badges that detect when they are talking, who they are close to and when they are moving about.
The workers were designing server systems. Over the course of a month, the researchers collected data on 911 individual jobs done by 23 employees in 1,900 hours. As well as measuring the time spent on each job — anything from five minutes to several days — they were able to control for its complexity and detect errors.
People who spent lots of time between jobs interacting with their colleagues — going to lunch or stopping for a chat — ultimately got much more done, the results showed. The best connected employee was 60% more productive than the least, says Waber, who presented his results at the International Conference on Network Science in Norwich, UK, on 27 June.
No one suspected that such interaction would help, says Waber. "The company was astounded — formally, these people were not supposed to be talking to each other," he says.
On the other hand, if someone communicated while they were assigned to a task — whether seeking help or distracted by others — their productivity dropped sharply.
The most productive workers showed the least variable behaviour. Waber thinks that such consistency might reflect what psychologists call a 'flow state', otherwise known as being 'in the zone'.
"Some consultants say you should talk to more people about what you're working on, but that might not be the case," says Waber. "Consistent behaviour might be more important while you're on a job."
Time to reflect
Such studies can be as surprising to the people actually doing the work as for those managing them, says sociologist Brian Uzzi of Northwestern University in Evanston, Illinois. "People are often oblivious to their own behaviour. They're blown away when they find out how they distribute their time."
But, he says, what works for one team might undermine another. "Tightly knit networks are really good for production teams that need to pool resources and share information, but they're terrible for marketing and distribution, because they don't have links to the outside."
Waber says that the company was "really encouraged" by the results and has started applying the study's lessons in its training and recruitment policies.
Does this mean that our employers will soon be monitoring our every cough and twitch in the name of productivity?
In this study, the data were anonymized, but there's no guarantee that will always be so. Transparency might be the answer, Waber suggests, so that employees can monitor their managers just as much as they are monitored from above.
Uzzi thinks that something similar to informed consent in medicine will become the norm, so that people joining a company can agree to how their data will be used. But for that to work, he adds, we will have to learn to read the agreement rather than blindly clicking the 'accept' box like we do for every website and piece of software we encounter.
Related links in Nature Research
About this article
Cite this article
Whitfield, J. Get a (social) life. Nature (2008). https://doi.org/10.1038/news.2008.925