When distributed working is not working
Sometimes the problem is the model rather than the execution, and it is worth being able to tell the difference.
Most problems attributed to remote work are execution problems: unclear expectations, poor documentation, meeting sprawl, a manager who has not adapted. Fixing the model would not fix them and would be a very expensive way to avoid the real work. For broader context on privacy and workplace data, ICO employment guidance is a useful reference.
But not always. Some situations genuinely suit distribution poorly, and recognising them is not a failure of commitment.
Symptoms that indicate execution, not model
- Decisions taking a long time — usually an unclear decision owner rather than distance.
- People feeling out of the loop — a documentation and routing problem.
- Meetings consuming the week — an audit problem.
- Managers feeling out of control — a defined-outcomes problem.
Each of these has a specific fix that does not involve anybody moving. Reaching for a return to the office to solve them is treating a process failure with a real estate decision. For a practical software-side reference on how assumptions can distort workplace judgments, this reference adds useful detail.
Situations that genuinely strain the model
Work requiring dense, continuous, high-bandwidth collaboration over long periods — early-stage product definition with a lot of ambiguity, crisis response, some kinds of creative work. Not impossible remotely, and materially harder.
Teams composed largely of people early in their careers, who learn substantially by observation and incidental correction. This is a real effect and it needs deliberate compensation rather than denial.
Work that is physically located: anything requiring hardware, a laboratory, a site or specific equipment.
Most of these are harder, not impossible, and the compensations are known. The failure is doing them without compensation and concluding that remote work does not function.
Ask people, structurally
Individual experience varies enormously and aggregate satisfaction hides it. Someone with a dedicated room and a settled home life has a different experience from someone working from a bedroom in shared accommodation.
Asking specifically — about workspace, isolation, career progression and whether they feel informed — produces actionable answers where a general satisfaction score produces none.
Partial changes before total ones
Between fully distributed and fully co-located there is a large space: periodic gatherings, a shared week each quarter, co-located sub-teams, an office available for those who want one.
These address most of the genuine difficulties at a fraction of the cost, and they preserve the hiring reach that was presumably the point. A total reversal should be a considered decision rather than a response to problems that have specific and much cheaper fixes.