The four-day week arrived with an argument attached. It was going to prove something about knowledge work, about how much of the week was theatre, about whether output and hours had ever been related. The trials were run as experiments with published results, and a great many of them reported success.
What survived is smaller and more useful than the argument. Companies still running a compressed week are mostly not running it everywhere, and they have stopped describing it as a position. It is a schedule, applied where it works, and the interesting question is where that turns out to be.
Coverage is the variable that decides
The pattern is consistent enough to state plainly. Where output is project-shaped — engineering, design, analysis, most of finance — the compressed week holds up, because the work is measured in completion rather than in availability and can absorb a day of nobody being reachable. Teams report the same throughput on four days, and the mechanism is rarely heroic effort. It is meetings.
Where the work is coverage — support desks, clinical care, operations, anything a customer or a patient arrives at unpredictably — it does not hold, and no amount of enthusiasm makes it. Reducing days without reducing hours of coverage means hiring, which is the cost the trial was meant to avoid. Companies that tried to apply one policy across both kinds of work generally ended up withdrawing it from everyone, having learned something about their own functions rather than about the four-day week.
That discovery is the durable finding. Teams that cut a day did it by killing recurring meetings, shortening the rest, and moving status reporting into writing, and the output held because those hours had not been producing anything. This is a finding about attention as the scarce resource nobody budgets rather than about the calendar, and it is available to any team willing to act on it without changing the schedule at all.
There is a management cost that the early enthusiasm understated. Compressing the week concentrates coordination into fewer hours and pushes more decisions onto individuals working without immediate access to a manager, which is fine on a team with room to breathe and difficult on one where the span of control has already been stretched to a number nobody defends. Several companies that withdrew the policy did so for this reason rather than for output — supervision, not productivity.
It also depends on a layer many firms removed. Running a four-day week well requires someone who can see the work clearly enough to know which coordination is load-bearing, which is the job of the middle managers companies cut and are quietly rebuilding. Without that judgement the meetings come back within two quarters, and the fifth day comes back with them.
The honest summary is that the four-day week works where the work is already legible, and the places it failed were mostly places that could not describe what their people did all day.
Topics leadershipproductivity
