You might think cutting the workday sounds radical, but plenty of companies have tested it and kept the change.
The idea is simple. Instead of paying for attendance, you pay for finished work and give people fewer hours in which to produce it. Long hours became a badge of honor in many startups, yet they feed the exhaustion that the World Health Organization classifies as an occupational phenomenon rather than a medical condition.
The real question is whether shorter hours hold up once the novelty wears off. This guide covers what the trials measured, what a shorter day costs, and how to test one without breaking service.
Key Takeaways
- A shorter workday means fewer paid hours per day at unchanged pay, not 40 hours squeezed into four longer days.
- The largest controlled trial so far, published in Nature Human Behaviour in July 2025, found less burnout and better health when hours fell without a pay cut.
- The recovered time comes out of meetings, interruptions and low-value messages, not out of people working faster.
- Shorter hours cost real money where coverage is continuous. Sweden’s best known trial showed the health benefits and the price tag together.
- Pilot with one team, define what “done” means before cutting hours, and measure output rather than time at a desk.
What a shorter workday actually means
Three quite different things get filed under “working less”, and they produce different results.
Fewer hours per day at the same pay
This is the genuine reduction: a six-hour or five-hour day across a normal week, with salaries untouched. Sweden’s Svartedalens care home and the US retailer Tower Paddle Boards are the examples people cite most, and both appear below.
A compressed schedule
Four ten-hour days still add up to 40 hours. Staff gain a day off, but every working day gets longer and more tiring. That can help with commuting and childcare, and it belongs in any discussion of flexible work schedules, but it is not a reduction in work time.
A shorter week
The 32-hour, four-day model removes a whole day instead of shortening each one. It is by far the best studied version, and we cover it in our guide to the four-day workweek.
Which model fits depends on the work. Roles built around short tasks, such as support, admin or scheduling, adapt well to shorter days. Project work needing long uninterrupted thinking blocks suits a shorter week better, because a five-hour day can be too short to get into a hard problem.
Why an eight-hour day rarely holds eight hours of work
The argument for shorter hours does not depend on people typing faster. It depends on how much of a standard day is already lost.
Microsoft measured this inside Microsoft 365 for its June 2025 Work Trend Index report on the “infinite workday”. The average worker is interrupted roughly every two minutes during core hours, and receives about 117 emails a day plus 153 Teams messages on a normal weekday. Almost six in ten meetings are ad hoc calls with no invite. Just under half of employees describe their working day as fragmented and chaotic.
Every interruption also has a tail. Research by Gloria Mark at the University of California, Irvine, widely cited since 2005, found people need well over 20 minutes on average to return to the original task. A day with a dozen such breaks is not an eight-hour day of work, whatever the timesheet says. Our piece on the modern attention span explains why that recovery time is so hard to shorten.
Cut the hours and something has to give. The first things to go are the standing status meeting, the reply-all thread and the third check-in of the day. A shorter workday is largely a meeting audit wearing a schedule as a disguise.
What the evidence shows
The largest controlled trial
In July 2025, Nature Human Behaviour published the biggest study yet of work-time reduction that keeps pay intact. Researchers followed 2,896 employees at 141 organizations across six countries over six months, and compared them with 12 control companies that changed nothing.
Employees at the trial companies reported lower burnout, higher job satisfaction and better mental and physical health. The control companies showed no such pattern. Three factors explained most of the improvement: people felt more able to do their job, slept better and were less fatigued.
One finding deserves attention: the bigger an individual’s own drop in hours, the bigger their gain. A policy that shortens the day on paper while some people quietly keep the old hours delivers little.
A national pilot with employer numbers
The UK’s 4 Day Week Foundation ran a six-month pilot from November 2024 to May 2025 with 17 companies and nearly 1,000 staff. Every company kept a shorter schedule afterwards: 12 moved to a four-day week, five to a nine-day fortnight. Among employees, 62% reported lower burnout, 45% greater life satisfaction and 41% better mental health. Sick leave fell and retention improved, which is why shorter hours now feature in serious conversations about talent retention.
The trial that showed the bill
Gothenburg’s Svartedalens care home ran a two-year six-hour-day experiment with 68 nurses on full pay. Staff were healthier, sick leave dropped and activities for residents increased. To keep the wards covered, the city hired 17 extra staff at roughly 12 million kronor, then chose not to extend the scheme, calling a general cut in hours too expensive to roll out quickly.
That case separates two questions that often get muddled. Does a shorter day improve health and output per hour? The evidence points that way. Does it pay for itself everywhere? Clearly not.
Tower Paddle Boards is the private-sector counterpart: founder Stephan Aarstol moved the company to a five-hour day and wrote a book about it in 2016. Treat it as one founder’s account, not measured evidence.
Where shorter hours run into trouble
Coverage-bound work
If a bed, a phone line or a machine must be staffed continuously, cutting hours means hiring. Healthcare, logistics, security and frontline support are the obvious cases, where staggered shifts or rotating teams beat a uniform cut.
Hidden overtime
The most common failure is a quiet one: the workload stays the same and people finish it after hours. Cap the workload, not just the calendar. Treat evening logins and weekend messages as a design fault rather than dedication. The debate around the right to disconnect matters here, because several countries now regulate exactly this.
Customers, policy and pay
Tell customers before you change, not afterwards, and give urgent requests a named route through. Most complaints come from surprise, not from the shorter hours. Internally, holiday accrual, overtime thresholds and benefits eligibility are written in hours, so bring payroll and legal in early.
Thinner culture
Less overlap means fewer unplanned conversations, and those carry a lot of informal learning. Replace some of it on purpose with short, regular rituals.
How to test a shorter workday
Start with one team and a fixed window
Choose a team whose output you can already measure and run for at least three months. A two-week experiment mostly measures novelty.
Define “done” before you cut hours
Write down the weekly outcomes each role owes and who signs them off. Shorter hours only work when priorities are explicit, which is why this pairs with outcome-based work.
Cut the meeting load first
Do this a month before the schedule changes, so you can see what the meetings were costing. Give every meeting an owner, an agenda and a 25-minute default, and move status updates into writing. Our guides to effective meetings and asynchronous work cover it.
Protect focused blocks and cut the noise
Reserve daily no-meeting hours and defend them from the top; a shorter day with the same interruption rate is just a squeezed day. See our guides to focus time, deep work and focus sprints. Then handle email in fixed slots (email batching), manage work notifications deliberately, and encourage single-tasking during core hours.
Automate the repeat work
Recurring reports, scheduling, data entry and first-line replies are the usual candidates. Every hour automation removes is an hour you do not have to find elsewhere; current business automation trends show how far that reaches.
Measure four things, then decide in the open
Track throughput, quality or error rates, customer response times, and how strained people say they feel. Hours worked is an input and tells you nothing. Publish the results, including anything that got worse, and say whether you are keeping, adjusting or stopping the model. A pilot that quietly fades out damages trust more than one that ends with reasons.
Conclusion
A shorter workday is not a productivity trick. It is a forcing function: cut the available hours and you have to decide what actually matters, a decision most teams postpone indefinitely.
The evidence supports the health case. Burnout, sleep and fatigue improve when hours genuinely fall and pay does not. The evidence on cost is mixed, and Svartedalens shows why: in coverage-bound work, somebody has to be paid to fill the gap.
So treat it as an experiment with a budget and an end date. Pick one team, write down what finished work looks like, cut the meetings before the hours, and measure output rather than attendance. If it works, you gain a healthier team and a clearer view of what the business needs done. If not, you learn where your real constraints sit. For the wider picture, see our overview of burnout prevention and of work-life balance best practices.
Found this useful?
Make SmartKeys a preferred source on Google, and our articles will surface more often in your Top Stories, AI Overviews, and AI Mode.
Add as Preferred Source
