Originally published 2020-05-12.

In August 2019, Microsoft Japan closed its offices every Friday and gave its roughly 2,300 full-time employees the day off on full pay. When the company compared the month with August 2018, sales per employee were up 39.9%, electricity use was down 23.1% and the number of printed pages had fallen by 58.7%. Of the staff surveyed, 92.1% said they liked the shorter week.
Those numbers made headlines around the world as proof that a four-day week boosts productivity by 40%. The result is genuinely interesting, but it is worth understanding what was measured, what else changed at the same time, and how it fits with other trials before drawing big conclusions.
What Microsoft Japan actually tried
The experiment was part of a summer programme called the Work-Life Choice Challenge. For five consecutive Fridays in August 2019, the offices were shut and employees took paid special leave. Pay was not cut.
The company did not simply remove a day and hope for the best. It also changed how people worked on the remaining four days:
- Meetings were capped at 30 minutes instead of the usual hour.
- Meetings were limited to five people.
- Staff were encouraged to use online chat and remote conferencing instead of gathering in person.
The programme also included initiatives around self-development and family wellbeing, which Microsoft Japan president Takuya Hirano presented as part of a broader push to let employees work and rest in ways that suited them. Some early reports mentioned a specific cash subsidy for family holidays, but the main English-language coverage of the results does not confirm an amount.
Watch: Here are the stories we're keeping our eye on: the first transcripts from closed-door depositions in the impeachment inquiry were released and a 4-day workweek increased a company's productivity. More news at https://t.co/9FZYNDqhCG. pic.twitter.com/cCE3bQtopP
— CBS4 Miami (@CBSMiami) November 5, 2019
What the 40% figure really means
The headline number refers to sales per employee compared with the same month a year earlier. That is a reasonable business measure, but it is not the same as measuring how much each person produced, or how good their work was.
Several things could have pushed the figure up besides the extra day off:
- Year-on-year comparison. Sales in a single month can rise or fall for reasons unrelated to working hours, such as product launches, large contracts or the wider economy. In 2019, many Japanese businesses were upgrading computers ahead of the end of Windows 7 support in January 2020, which may have boosted demand.
- Several changes at once. Shorter meetings and fewer attendees may have saved a lot of time on their own. The trial does not separate their effect from the effect of the free Friday.
- A short window. One month is long enough to show enthusiasm, but not long enough to show whether gains last once the novelty wears off.
The savings on electricity and paper are more straightforward. Closed offices use less power, and fewer meetings mean fewer printed handouts. Even so, those figures also cover only one month.
None of this means the trial failed. Employees were clearly happy with it, and the company planned a further winter challenge to encourage flexible working. It simply means the 40% figure is a promising signal from one company, not a law of nature.
Why Japan was a striking place to test it
The experiment drew extra attention because of where it happened. Japan has long been known for very long working hours and a culture in which leaving the office before your manager can feel awkward. The country even has a word, karoshi, for death from overwork, and the government introduced reforms around 2019 to cap overtime.
A large, well-known employer in that setting openly testing fewer days, and reporting good results, was a powerful message. It suggested that in some workplaces the length of the week may matter less than how the time is used.
Other four-day week trials
Microsoft Japan was not the first organisation to try this, and it was far from the last.
Perpetual Guardian, New Zealand (2018). The estate-planning firm, which had around 240 staff, let each employee take one day off a week on full pay during a trial in March and April 2018. The company reported higher productivity, lower stress and better work-life balance, with revenue holding steady while costs fell. It later made the policy permanent, and its founder, Andrew Barnes, went on to co-found the campaign group 4 Day Week Global. Some staff did say that the shorter week increased pressure around deadlines.
Iceland (2015 to 2019). Trials involving about 2,500 public-sector workers cut hours to 35 or 36 a week without reducing pay. Wellbeing improved and productivity stayed the same or improved in most workplaces. Strictly, most of these were shorter weeks rather than four-day weeks, as many workplaces trimmed up to three hours rather than a full day.
United Kingdom (2022). A six-month pilot organised by 4 Day Week Global involved 61 companies and around 2,900 employees. Most firms chose to continue afterwards, and 18 made the change permanent. Staff turnover fell significantly.
Not every experiment has been a clear success. Utah moved state employees to four ten-hour days in 2008 to save money and dropped the scheme in 2011, and some studies of four-day school weeks in the United States have linked them to lower test scores. The format seems to work best when it is designed around the job, rather than imposed as a cost-cutting measure.
What the research says about shorter days
Older versions of this story cited a 2018 Harvard Business Review "study" showing that six-hour days make people more productive. That piece, The Case for the 6-Hour Workday by entrepreneur Steve Glaveski, is an opinion essay rather than research. It makes a thoughtful argument that knowledge work depends on uninterrupted focus, and it recommends shorter meetings and fewer notifications, but it does not present new data.
A 2018 survey by the Workforce Institute at Kronos and Future Workplace, covering nearly 3,000 employees in eight countries, found that most respondents saw four days or fewer as their ideal working week, and nearly half believed they could do their job in five hours a day if interruptions were removed. That tells us how workers feel, which matters, but it is self-reported rather than a measure of output.
The stronger evidence comes from the organised trials above, which repeatedly find better wellbeing and stable or improved output. The common thread is that the time saved has to come from somewhere: fewer and shorter meetings, less pointless email, and clearer priorities.
If your workplace wants to try it
For anyone considering a similar experiment, the Microsoft Japan story offers some practical lessons:
- Decide what success means before you start. Pick a few measures, such as output, customer satisfaction, error rates and staff wellbeing, and record a baseline.
- Change the work, not just the calendar. Cut meeting lengths, trim attendee lists and protect time for focused work.
- Run it long enough. Several months will tell you more than a single month.
- Plan for customers and cover. Staggered days off can keep services running.
- Listen to staff. Some people find a compressed week stressful; flexibility matters.
The honest takeaway from Microsoft Japan's trial is not that everyone will be 40% more productive with Fridays off. It is that many workplaces waste more time than they realise, and that giving people a reason to use their hours well can benefit both employees and the business.
Sources: CNBC: Microsoft Japan four-day week experiment (November 2019); World Economic Forum: Microsoft Japan's four-day week; Wikipedia: Four-day week; Harvard Business Review: The Case for the 6-Hour Workday (2018).

COMMENTS