The four-day workweek stopped being a thought experiment several years ago — it now has multiple published, multi-year pilots with real retention, revenue, and burnout data behind it. The dominant model tested is "100-80-100": 100% of pay, 80% of the time, for 100% of the output. Across the UK's large-scale 2022 trial, Iceland's earlier public-sector trials, and Microsoft Japan's one-month experiment, the pattern holds up more consistently than most workplace pilots ever do: companies that switch rarely go back, and the productivity hit most people expect largely does not materialize.
What changed in 2026
- More countries moved from pilot to policy. Following earlier trials, several European countries introduced a formal right for employees to request a compressed schedule, shifting the conversation from "would this work" to "how do we implement it."
- Multi-year follow-up data is now available. The UK's 2022 pilot reported roughly nine in ten participating companies still running a four-day week a year or more later, a stronger signal than the initial trial numbers alone.
- AI-assisted workflows changed the output math. Companies citing AI tools for drafting, scheduling, and first-pass analysis report the harder question is no longer "can we fit five days into four" but "how much of the fifth day was never necessary."
- The conversation split sharply by industry. Adoption remains concentrated in knowledge work, professional services, and tech; manufacturing, healthcare, and retail have had far more trouble fitting the model around shift-based staffing.
What the pilots actually found
| Pilot |
Scale |
Headline result |
| UK trial (2022) |
61 companies, ~2,900 employees |
Revenue held roughly flat or grew slightly; most companies kept the schedule afterward |
| Iceland trials (2015-2019) |
~2,500 workers, about 1% of the workforce |
Deemed a success by researchers; led to widespread reduced-hour agreements afterward |
| Microsoft Japan (2019) |
Single company, one month |
Reported a large productivity jump alongside lower electricity and printing costs |
| Later multi-country trials |
Dozens of companies across several countries |
Broadly consistent: burnout and stress down, resignation intent down, output flat to higher |
The consistent thread across all of them is not that people worked harder in four days — it is that a large share of the fifth day in a normal week was low-value meetings, context-switching, and busywork that simply got cut rather than compressed.
How to tell if your company could actually do it
- Audit where the current fifth day actually goes. Track a real week and separate deep work from meetings, admin, and idle time. If a large share is already low-value, compression is realistic; if the fifth day is already lean, it is a harder sell.
- Pick a model that matches your work type. A uniform day off works for teams with predictable, individual-heavy work. Staggered days off work better for teams that need daily coverage.
- Run a real trial with a fixed length. Three to six months, with output metrics agreed in advance, not vibes. The pilots that succeeded measured specific things before comparing after.
- Protect meeting discipline hardest of all. Every pilot that succeeded paired the shorter week with aggressive meeting cuts — fewer, shorter, and more tightly agenda'd. Without this, four days becomes five days of work compressed into four, which fails fast.
- Decide the failure criteria before you start. Agree in advance what would cause you to revert, so the trial is judged on the metric, not on anecdote after one rough week.
Common mistakes
- Compressing five days of work into four instead of cutting the work. This produces the burnout the model is supposed to fix. The point is removing low-value work, not fitting it into a smaller box.
- Skipping the meeting audit. Nearly every failed internal trial kept the same meeting load. Meetings are usually the first thing that has to shrink for the math to work.
- Choosing the wrong model for the work type. A uniform day off does not work for a team that needs five-day coverage; a staggered model does not suit heavily collaborative teams. Match the model to the actual work pattern.
- Running the trial without a measurement plan. Without agreed metrics up front, the pilot gets judged by whoever had the worst week, not by the data.
FAQ
Does a four-day week actually hurt productivity?
The published pilots mostly report output holding flat or improving, largely because low-value work gets cut rather than compressed into the remaining four days.
Does this work for every industry?
No — it is easiest in knowledge work and professional services, and hardest in shift-based, coverage-dependent industries like healthcare, retail, and manufacturing, where a fifth day of staffing usually cannot simply disappear.
Is this the same as a compressed schedule of 10-hour days over 4 days?
No — the pilots discussed here test the 100-80-100 model: full pay for reduced hours, not the same hours compressed into fewer, longer days.
How long should a company trial it before deciding?
Three to six months is the common range in successful pilots — long enough to get past the novelty effect and see a real output trend.
Where to go next
If your company is not ready for four days, hybrid work best practices covers a more incremental flexibility model. For the opposite pressure, see return-to-office negotiation tips, and pair either with slow productivity for the individual-level version of doing less, better.