Remote working led to drop in high quality output, Microsoft study concludes

A new study of 61,000 Microsoft employees claims that the sudden shift to remote working had a profound effect on their work behaviours and output. Overall, remote workers spent less time in meetings, had fewer real time conversations and worked in more siloed ways. Crucially it found that there was a reduction in the strengths of ties between people and fewer networking opportunities. While not suggesting that office based work and remote work are necessarily better than the other, the study concludes that the drop in loose connections and chances of networking with other teams could have a negative impact on higher quality work outputs and working culture.

The study was published earlier this month in the journal Nature Human Behaviour and co-authored by Berkeley Haas Assistant Professor David Holtz Holtz while working as an MIT Sloan doctoral intern at Microsoft alongside Microsoft colleagues.

The report says that prior to the COVID-19, pandemic no more than 5 percent of Americans worked from home more than three days per week, whereas it is estimated that by April 2020 as many as 37 percent of Americans were working from home full-time. The study explored the impact of sudden remote working at this scale and also the effects over the longer term.

Examining the working patterns of Microsoft employees, including those who had previously worked remotely, the study set out to distinguish between changes solely or largely due to remote work and those related to the extraordinary circumstances in which people were living. The inclusion of Microsoft employees who had previously worked remotely allowed the researchers to take account of external factors related to COVID.

The study found that the shift to remote work caused both formal and informal communities within Microsoft to become less interconnected and more siloed. Remote work caused the share of collaboration time employees spent with cross-group connections to drop by about 25 percent of the pre-pandemic level.

Furthermore, firm-wide remote working caused separate groups to become more ‘intraconnected’ by adding more connections within themselves. The shift to remote work also caused the organisational structure at Microsoft to become less dynamic; Microsoft employees added fewer new collaborators and shed fewer existing ones.

“Microsoft employees didn’t just change who they worked with, but also how they worked with them,” the authors say. “The results indicate that the shift to firm-wide remote work increased unscheduled call hours but decreased total meetings and call hours by 5 percent of their pre-pandemic level. This suggests that the increase in meetings many experienced during the pandemic was not due to remote work, but due to the pandemic and related factors.

“Remote work also increased asynchronous communication, like email and IM. Based on previous research, we believe that the shift to less ‘rich’ communication media may have made it more difficult for workers to convey and process complex information.

“Our methodology also allowed us to decompose the overall effect of firm-wide remote work into two separate components: the effects of someone working remotely on their work practices, and the effects of someone working remotely on their collaborators’ work practices. We found that many of the effects of firm-wide remote work were driven as much by effects on collaborators as effects on oneself. This suggests that even if firms were to adopt work arrangements where some employees return to the office, the effects on collaboration patterns would be still be present, albeit somewhat mitigated.

“In light of these findings, companies will need to take proactive measures to try to help workers acquire and share new information across groups, so that productivity and innovation are not impacted. Separating the effects of working from home from the effects of COVID-19 requires statistical knowhow. Furthermore, the effects of these policies on company culture and innovation could take years to measure. Because remote and hybrid work are likely to persist even after the pandemic has ended, it is incredibly important to understand how these policies affect the ways that people collaborate with one another.”