Open offices are as bad as they seem—they reduce face-to-face time by 70%

Tearing down walls and cubicles in offices may actually build up more barriers to productivity and collaboration, according to a new study.

Employees at two Fortune 500 multinational companies saw face-to-face interaction time drop by about 70 percent, the use of email increase between 22 percent and 56 percent, and productivity slip after their traditional office spaces were converted to open floor plans—that is, ones without walls or cubicles that ostensibly create barriers to interaction. The findings, published recently in the Philosophical Transactions of the Royal Society B, suggest that removing physical dividers may, in fact, make it harder for employers to foster collaboration and collective intelligence among their employees.

Many companies have waged a so-called “war on walls” to try to create such vibrant workspaces, the authors Ethan Bernstein and Stephen Turban of Harvard wrote. But, “what they often get—as captured by a steady stream of news articles professing the death of the open office—is an open expanse of proximal employees choosing to isolate themselves as best they can (e.g. by wearing large headphones) while appearing to be as busy as possible (since everyone can see them).”

Before the study it was clear from employee surveys and media reports that workers are not fans of the open architecture trend. Employees complain of noise, distractions, lowered productivity, a loss of privacy, and a feeling of being “watched.” On top of that, studies have suggested that open offices can be bad for workers’ health.

Still, Bernstein and Turban write that, up until now, there has been a dearth of data on how employee behaviors change in these boundless, despised work spaces. To come up with that data, they enlisted employees in two big companies as their employers embarked on remodeling office areas from traditional closed offices and cubicles to open, boundary-less space.

In the first anonymous company—code-named OpenCo1—the researchers tracked the activity of 52 employees for 15 days prior to the redesign and then, three months later, another 15 days after the redesign. The three-month gap allowed the employees—who worked in sales, human resources, technology, product development, and leadership—to settle into their new work environment.

The employees wore a badge with a set of sensors that monitored their interactions and movements: an infrared sensor captured who they were facing; a microphone picked up office chatter; an accelerometer recorded movement; and a Bluetooth sensor tracked their location in the office. A face-to-face interaction was recorded when employees' infrared sensors faced each other, microphones picked up alternating speech, and location monitors indicated employees were close to each other. The interaction ended when any of these conditions ceased for five seconds. The researchers also monitored electronic communications between employees.

In the end, the researchers captured 96,778 face-to-face interactions, 84,026 emails, and 25,691 instant messages.
Out of bounds

After the open redesign, the employees spent 72-percent less time having face-to-face interactions with each other. The raw numbers shook out to an average of 5.8 hours of face-to-face time per day per person before the redesign, but only 1.7 hours of face-to-face time per person per day afterward. Meanwhile, electronic communication increased; workers sent 56 percent more emails to their colleagues and 67 percent more instant messages.

Executives for the company also shared in qualitative terms that productivity among employees—measured by internal, confidential metrics—had also fallen. They did not disclose the magnitude of the productivity decline.

The researchers then, essentially, repeated this observational-type study in the second anonymous company, OpenCo2. But this time, they enrolled more employees and monitored them for longer time periods. That is, they tracked 100 employees for eight weeks before a similar office overhaul and two months afterward for another eight weeks. The researchers also collected information on pairs of people known to interact with each other—dyads—as well as employee attributes, including gender, role, and desk location.

The researchers found very similar effects to those seen in OpenCo1. Face-to-face interaction time declined between 67 percent and 71 percent in OpenCo2, depending on how the researchers parsed the data on dyads and desk distance. And emailing jumped between 22 and 50 percent. (The researchers didn’t look at instant messaging data for this company.) Unsurprisingly, workers with desks close together or who worked on the same team had more interactions—but the effects were smaller than expected. Gender was not a factor in interaction levels.

Together, the researchers conclude that the wall busting backfired. They speculate that the reasons may include the need for workers to have constraints on their interactions and have boundaries that help them make sense of their environment. Privacy concerns may also be a key reason. They write:

Consistent with the fundamental human desire for privacy and prior evidence that privacy may increase productivity, when office architecture makes everyone more observable or ‘transparent’, it can dampen [face-to-face] interaction, as employees find other strategies to preserve their privacy; for example, by choosing a different channel through which to communicate.
But perhaps more notably, they point out that researchers and employers have yet to come up with a breakthrough leading to that coveted, buzzy work environment that amplifies creativity and collective intelligence. It’s just clear that breaking down actual walls isn’t the way to get there.