Every 4-day pilot meets the same quiet saboteur: the employee who takes Friday “off” and works it anyway. They answer the urgent email, finish the deck, clear the inbox so Monday feels survivable. Individually it looks like diligence. Across a pilot, it corrupts your data, exhausts your most conscientious people first, and teaches everyone else that the day off is not really real.
Why conscientious staff work their day off
The drivers are rarely laziness about boundaries. They are guilt (teammates covering for me are struggling), fear (my workload will be judged on five-day expectations), and backlog logic (an hour on Friday saves three on Monday). Managers create a second driver without meaning to: praising the person who “heroically” cleared the queue over the weekend, or scheduling Monday deadlines that quietly assume Friday labour.
There is also a design driver. If a team’s work was not redesigned before launch — no stop-doing list, no handover discipline, no cover rota — then Friday’s work still exists, and someone absorbs it. The most committed person absorbs it invisibly. That is why secret working concentrates in exactly the staff you can least afford to burn out.
What it does to your pilot
A pilot that “succeeds” on hidden overtime has not succeeded; it has mis-measured. Your output figures include Friday hours your cost model no longer contains, your wellbeing scores flatter a benefit people did not fully receive, and your decision at month six rests on numbers that collapse the moment the secret workers stop, get sick or leave.
The UK 2022 pilot is a useful calibration. Average hours fell from about 38 to about 34 a week — a real reduction, but not the full 32 the headline implied. The lesson is not that firms cheated; it is that realised reductions are usually smaller than policy reductions, and honest pilots measure the realised figure. Portugal’s 2023 pilot added a second warning from the other direction: continuation tracked whether firms changed how work was done, not whether they announced the day. Where process change was thin, the model thinned with it.
How to see it without spying
You do not need surveillance. You need aggregates and conversations. System-level counts — logins, messages sent, documents edited on days off — reported at team level, never as individual naming, show the shape of the problem within a week. After-hours totals belong in the same dashboard as output, so a team hitting targets with rising Friday activity triggers a workload conversation rather than praise.
Managers need a script for that conversation, because “stop working Fridays” is not one. It is: what landed on you this week, what should have waited, whose cover failed, and what do we stop, delegate or re-time so next Friday is actually off? Ask it monthly and record the answers; the list of recurring Friday tasks is your redesign backlog, gift-wrapped.
Then fix the signals. Leaders who send nothing on the day off. Monday deadlines set with four-day arithmetic. Cover that routes to a named buddy instead of back to the absent person’s inbox. And public credit for the team that hits its targets with Fridays genuinely quiet — that is the behaviour the model runs on.
The conversation to have at launch
Say the quiet part in the launch meeting: some of you will be tempted to work the day, especially in month one; that will feel helpful and will actually harm the pilot; if your workload cannot fit four days, that is information we need, raised in the open, where we can redesign it. Pilots that name secret working early report it honestly. Pilots that rely on everyone simply behaving discover it late, in the attrition figures.
Practical takeaways
- Treat day-off activity as workload data, not misconduct — investigate the work, not the worker.
- Track team-level system activity on days off alongside output, from week one.
- Give managers a monthly script: what landed, what should have waited, what will we stop?
- Fix the signals: nothing sent by leaders on the day off, deadlines set in four-day arithmetic, cover to a named buddy.
- Name the temptation at launch and invite staff to surface overload early rather than absorbing it silently.
Sources
- Autonomy, University of Cambridge and Boston College, UK four-day week pilot results, reported February 2023 — https://www.sciencedaily.com/releases/2023/02/230221113132.htm
- Portugal four-day week pilot, Birkbeck and Henley Business School, reported by Personnel Today, 14 December 2023 — https://www.personneltoday.com/hr/portugals-four-day-week-reduces-anxiety-and-fatigue/
- Autonomy, UK pilot one-year follow-up, reported February 2024 — https://www.hrdconnect.com/2024/02/26/one-year-on-uks-4-day-work-week-trial-has-been-a-success/
Day5Group is a consultancy that assists in the transition to a 4 day work week. More about Day5Group · All blog articles