A new study by Ethan S. Bernstein, an associate professor in the Organizational Behavior unit at Harvard Business School, shows that open plan offices actually result in a sharp decrease of face-to-face interactions at work. While previous research and reports on the topic of open plan offices replied on subjective, self-reported data from employees in cubicle-free workplaces, the new study used sensor-filled "sociometric badges" to record and measure the daily activities of employees.
The badges collected information using an infrared sensor that tracked who each employee was interacting with. Additionally, researchers used microphones that detected if the wearer was talking or listening, accelerometers that measured movement and posture, and a Bluetooth sensor that determined spatial location inside the office—all recorded with a granularity of 10 milliseconds. Using these sensors, a face-to-face interaction was counted when two or more badges could make a line-of-sight connection, detected alternating speaking, and were recorded as being within 10 meters of each other. This data was compared with the email and messenger data recorded at the same time.
For the first of the two studies, conducted at unnamed Fortune 500 multinational companies, a group of employees across different roles were measured in their original office space for three weeks, and then again in a open plan office space with "similarly assigned seats on a redesigned floor of the same size," for another three weeks. The company in question was in the process of redesigning their offices as part of what the researchers characterized as a "war on walls" in order to increase face-to-face interaction.
SEE: Comparison chart: Enterprise collaboration tools (Tech Pro Research)
Contrary to this intended outcome, the participants in the study spent 72% less time in face-to-face interactions, with interaction decreasing from "roughly 5.8 [hours] of F2F interaction per person per day" to only 1.7 hours per day. Accordingly, the researchers found that participants sent 56% more emails, received 20% more emails, and were copied on 41% more emails from other participants, with messenger frequency increasing by 67%, and words sent by IM increasing 75%.
Likewise, for the second study, data was collected for eight weeks, and then again for eight more weeks, two months after the employees had moved to a new open plan office. Due to differences in how the second study was measured (taking into account gender, team assignment, role, and desk location,) the calculations were slightly different, though the results largely identical. The employees tracked in this study interacted between 67% and 71% less than before, with emails increasing between 22% and 50%. The researchers noted that dyads—conversational pairs—in the same team or role communicated more in face-to-face interactions and email compared to dyads on different teams or roles. Interestingly, short physical distance increased, slightly, the amount of face-to-face interaction, but not email. No significant change was measured in interaction by gender.
According to the study, there is a discrete difference between architectural changes and taking social measures to foster open interaction. Accordingly, extrapolating findings from research into swarm intelligence among social insects does not necessarily apply to human interaction. According to the researchers:
We see a close relationship between our finding that open, 'transparent' offices may be overstimulating and thus decrease organizational productivity... finitely bounded, and often small, group size maximizes decision accuracy in complex, realistic environments. Similarly, recent collective intelligence work suggests that, like our open offices, too much information from social data can be problematic, partly because of challenges focusing attention, but also for reasons that extend to more general functions of human cognition.
The researchers also noted that changes from privacy-friendly office plans, compared to open plan offices, also have an effect on the medium by which people communicate, but also to whom they seek communication with, which they claim "can have profound consequences for how—and how productively—work gets done."
The big takeaways for tech leaders:
- Researchers found that open plan offices reduce face-to-face interaction by roughly 70%, despite the intended result of increasing interaction by removing physical barriers.
- This study is seemingly the first such study to measure interactions using IoT sensor badges rather than relying on self-reported surveys about interaction and work satisfaction in open plan offices.
- How to optimize the smart office (ZDNet special report) | Download the report as a PDF (TechRepublic)
- 3 ways the 'smart office' will change the future of work (ZDNet)
- Slack: The smart person's guide (TechRepublic)
- 7 must-have devices for your smart home office (ZDNet)
- 63% of employees say open concept workplaces actually decrease productivity (TechRepublic)
James Sanders is a Tokyo-based programmer and technology journalist. Since 2013, he has been a regular contributor to TechRepublic and Tech Pro Research.