Remote Work Changes the Way Employees Communicate and Collaborate with One Another

In March 2020, as the pandemic was starting, Microsoft asked all its non-essential employees to shift from full-time office-work to 100% remote. Because this change was triggered by sudden and unexpected changes in circumstances (researchers call such events “exogenous shocks”), it allowed a team of researchers from Microsoft, Berkeley and MIT to estimate the causal effect of firm-wide remote work on employees’ communication and collaboration patterns.

Mar 6, 2024 Organizational issues

What are the results?

First, business groups within the company became more siloed. Cross-group interactions decreased: Employees spent less time interacting with people outside of their group, which reduced their ability to access new information located in other parts of the network. At the same time, in-group ties reinforced: Members of a team became more closely connected to one another, which made professional networks denser.

Second, the networks of employees became more static: People added and deleted fewer ties and spent less time with their newly added ties.

In brief, the shift to a fully remote work organization made employees’ networks more rigid (i.e., fewer changes over time), more fragmented (i.e., less communication across groups), and more clustered (i.e., each employee’s network became denser). Employees spent more time within their group and less time with colleagues outside of their group. However, it is typically through those weak ties between groups that people access to new information. These changes therefore had an impact on the information flow within the organization, by making access to new information more difficult.

The researchers also observed changes in communication methods. Since in-person communication was no longer possible, one might have expected that employees relied on other “synchronous” forms of communication such as phone calls, live meetings, or video calls. Instead, synchronous communication was replaced by asynchronous communication, in the form of emails and instant messages.

This, too, had an impact on the information flow: Because asynchronous communication is not as information rich as synchronous communication, it typically prevents the transfer of complex ideas. Knowledge transfer therefore becomes more difficult.

How did they show that?

One cannot determine the effect of remote work by comparing people who work remotely to those who don’t: Indeed, people who have access to remote work, and choose to work remotely, are probably different in many ways from the ones who work from the office. If we observed a difference between them, we could not conclude that it is uniquely due to remote work.

To determine the causal effect of remote work on employees’ network, researchers would typically need to run an experiment: Employees would be randomly assigned to a treatment group (work from home) or to a control group (work from the office), and researchers would study the difference between the two. However, an experiment like this would be extremely impractical to conduct: It would disrupt the life of the organization in unexpected ways.

Instead, the authors took advantage of a “natural experiment” in which an unexpected and random shock (the pandemic) forced everybody to transition to remote work. Since this shock affected everybody in equal ways, independently of their characteristics and decisions, researchers could then compare employees’ network before and after this shift.

More precisely, the team of researchers collected 6-month (from December 2019 to June 2020) of communication data (emails, calendars, instant messages, video/audio calls, work hours) from around 60,000 Microsoft employees and compared those data before and after the shift to a full-time work from home organization (in March 2020).

To make sure that any change that they observed was linked to the shift to remote work (as opposed to other things that the pandemic might have triggered, such as anxiety, or lack of childcare), they used a “control” group: People who were already working remotely before the pandemic (18% of the staff). Since they observe much smaller changes among these “already remote” employees, they can confidently attribute the observed changes to remote work (as opposed to the pandemic itself).

Why is it happening?

Authors provide few explanations to explain this change in employees’ network, but we can speculate.

Office work is full of serendipitous interactions: Colleagues bump into each other in the hallways, meet at the coffee machine, at the printer, or at the cafeteria; commute to work together. They see that their colleagues’ door is open and stop by to say hi. They take a few minutes before or after a meeting to discuss the meeting, or other things.

All these informal interactions give employees opportunities to ask questions, gather information, get tips, and meet people… In a remote work context, such informal and random interactions tend to disappear: To chat with someone, one must purposefully call them, or send an email. This might explain why people have fewer interactions with people outside of their team, and mainly interact with people with whom they work.

Other negative consequences of remote work

The changes observed in the communication and collaboration network are not the only consequences of remote work. When interviewing people around me about their remote work experience, several other things came up:

Virtual and asynchronous communication (which tend to dominate in a remote work context) can quickly lead to conflicts. In a remote work context, communication is impoverished: Most of the cues that normally minimize the odds of misunderstandings (body language, tone of voice, facial expressions…) become invisible. Misunderstandings between parties increase, people misinterpret signals, see innuendos where there are none… Minor issues are more likely to escalate into tensions, or even full-blown conflicts.
Tacit and informal rules are much harder to grasp in a virtual world, particularly for new members of the organization. The corporate culture, for instance, is rarely codified: It is learnt by observing how employees speak, move, and act within the organization. The same is true of power dynamics: One would rarely write in an email “you should stay away from this person, they will make your life difficult”, but you might learn this valuable information through informal chat at the coffee machine.
Asynchronous forms of communication can be exhausting. To iron out the details of a problem, or to make sure that everybody is on the same page, it can be much easier to drop by someone’s office and take five minutes to chat. Emails and instant messages can instead force people into frustrating back-and-forth exchanges and ultimately offer no resolution to the initial problem.
Is it hard to make friends in a remote work context! This point is often overlooked, because friendships at work are often perceived as a “nice-to-have”, but not something essential. In fact, the opposite is true: Research suggests that having friends at work is positively associated with performance, job satisfaction, well-being, feeling of belongingness, trust among co-workers… On the contrary, a lack of personal connections among employees can lead to lower level of trust, reduced cooperation, or higher turnover… The social bonds that employees create among themselves can be an asset for the company, but a fragile one that needs to be nurtured.

How can organizations avoid this tendency?

First, is it really an issue if employees have denser networks, develop fewer new ties, and spend less time with people outside of their team? Unfortunately, research on social networks within organizations suggests that it is indeed a problem.

Dense networks, with few weak ties and few bridging ties, are networks in which novel information does not flow, and therefore that hinder creativity and innovation. Indeed, good ideas are often generated by combining unusual, novel pieces of information together, and this information is often coming from various parts of the network. However, in a network in which everybody knows everybody else, and in which people talk to each other all the time, information is “redundant”: Everybody knows the same things.

Organizations therefore face a challenge: On the one hand, they must respond to employees’ demand for more flexibility and work-life balance. On the other, they must be mindful that remote work might not always be the most efficient for communication and collaboration.

Given how much employees’ priorities have changed, it is hard to imagine that organizations will simply come back to full in-person work. However, organizations need to tackle the specific challenges of remote work: How do you create a vibrant organizational culture, in which everybody feels included? How do you make sure that people “get out of their bubble”, and can learn from other people located somewhere else in the network? How do you minimize the conflicts that can arise when people rarely see each other face-to-face?

Some solutions might be:

To impose regular in-person gatherings.
To raise awareness about the potential issues that may arise in a remote work context: Patterns of communication might change, professional networks might shrink, access to new information might become difficult, asynchronous media of communication can be problematic for certain types of exchanges (e.g., transfer of complex, non-codified knowledge).
To train managers on how to anticipate and minimize conflicts in a remote work setting.
To emphasize the need of clarifying (at the risk of “over-explaining”) thoughts and intentions in virtual communication channels.
To explore the possibilities of a “virtual open-door policy” in which employees can hop on a call at any time when they face difficulties.