Remote Work and IS: What Actually Changed

In March 2020, a hospital system I read about in press coverage at the time sent approximately 4,000 non-clinical staff home with instructions to work remotely. They had no remote work policy. They had no standard laptop configuration for home use. They had a VPN designed for a few hundred remote users at a time. Within 48 hours, the VPN was effectively unusable.

That story happened everywhere, at different scales, across almost every industry. The COVID-19 pandemic forced mass adoption of remote work across organizations that had never planned for it. It was not an IS project with a requirements phase and a rollout plan. It was an emergency. The IS infrastructure either held or it did not. Most of it did not, at first.

What makes this interesting from an IS perspective is not the technical problems, which were predictable in retrospect. VPN capacity had been sized for a small percentage of users working remotely at any time. Organizations suddenly needed 100 percent remote capacity simultaneously, and they had not built for that. Video conferencing (Zoom, Teams, Meet) went from a supplementary tool to the primary means of organizational communication essentially overnight. Zoom's usage numbers, widely reported at the time, grew by an order of magnitude in a matter of weeks. The service held better than many expected, though there were stability issues in the early weeks.

The more interesting problems were organizational. Many organizations discovered that their processes depended on physical proximity in ways they had never articulated, and probably never thought about. The quick question to a colleague becomes an email or a message that might not get answered for hours. The hallway conversation where someone mentions a problem they are stuck on and a colleague offers an offhand suggestion that solves it does not translate cleanly to scheduled video calls. The whiteboard session, which is so embedded in how technical teams work that "whiteboard" became a verb, turns out to be hard to replicate with shared document tools when you are also trying to maintain eye contact with six faces in a grid.

Socialization and onboarding were perhaps the most clearly broken processes. A new employee who joins an organization remotely has to learn the culture, the informal relationships, the unwritten rules, and the technical environment all at once, without being able to observe anything directly. Organizations learned in 2020 and 2021 that their onboarding processes were not actually documented. They were performed by proximity.

Nick Bloom and colleagues at Stanford have been tracking remote and hybrid work patterns and productivity since before the pandemic, and their work became widely cited during and after the emergency. From what I have followed of their research, the broad finding has been that fully remote work tends to show modest effects on productivity for most knowledge work, with significant variation by role and task type, and that hybrid arrangements show more variable results depending on how they are structured, particularly whether hybrid means "some days in office" versus "rarely in office." I am describing this directionally because I do not have precise numbers in front of me that I have verified, and the findings have evolved as the research has continued.

The IS infrastructure dimension is one part of the story. The organizational dimension is another. The one that I think is underappreciated is the record-keeping transformation that happened almost accidentally.

Before the pandemic, a large portion of organizational decision-making happened in rooms. People gathered, discussed, decided, and dispersed. The record of those decisions, if any existed, was whatever notes someone took and wherever those notes went. Often they went nowhere. The decision lived in the memory of the people who were in the room. Sometimes only in the memory of the most senior person there.

In organizations that shifted to remote work using Teams, Slack, or similar platforms, that changed. Decisions that happen in a Teams chat have a record. That record is searchable. It has timestamps and participants. It can be read six months later, or three years later, or by a lawyer conducting discovery in a dispute. This is not a new capability in principle. Organizations have had email for decades. But the shift to using chat platforms for the kind of informal, real-time decision-making that used to happen in rooms pushed that decision-making into a medium with automatic archiving.

The organizational memory implications are substantial. A company with five years of Teams or Slack data has a record of a large portion of its operational decision-making. Whoever controls the search and interpretation of that archive has access to something that simply did not exist before in most organizations. This changes accountability (there is a record of who said what), knowledge management (institutional knowledge that used to walk out the door when someone left is partially captured), and legal exposure (discovery in litigation can now reach conversations that would previously have been unrecoverable).

Most IS infrastructure discussions about the pandemic focus on whether the tools worked technically and whether productivity held. Those are reasonable questions. But I think the record-keeping transformation is the deeper change, and most organizations have not fully absorbed what it means.