One of the concerns that some people raise about the 4-day workweek is that while it might work great at little places that are nimble and flexible, it’s not going to work at larger organizations that have a lot of diverse functions. Size isn’t actually the limit people think: as my book SHORTER (US | UK) explains, there are companies in Japan and Korea that have a thousand people or more and have moved to 4-day weeks or 6-hour days In recent years.

This summer, Microsoft Japan successfully trialed a 4-day workweek for its 2300 employees. During the month of August, the company closed on each of the five Fridays, then measured the impact on everything from electricity use and employee happiness, to paper use and employee productivity to the number of meetings the company held.

It should come as no surprise that the results were very positive.

There’s some background that isn’t always getting covered in the English-language press about the experiment that is worth noting.

First, the Japanese government and Japanese companies are conducting a lot of experiments now to find improved ways of working, and to cut back on the notorious culture of overwork. Partly this is driven by the aging of the workforce and other social concerns, and partly it’s an effort to promote remote work before the 2020 Olympics, in order to reduce congestion. The Metropolitan Tokyo government, for example, ran “Telework Days 2019” from late July to early September.

Second, the 4-day week challenge also builds on years of earlier Microsoft programs aimed at creating greater flexibility and work-life balance. Microsoft Japan started encouraging remote work in 2012; two years later their “Telecom Day” had become “Telecom Week,” and two years after that, “Working Style Reform Week.” In fact, since 2015, working hours had been reduced 80 hours per year (about 600K hours across the company). Like at Cybozu, they’re trying to broaden flexibility and choice, and accommodate a variety of different family schedules. Microsoft Japan had also been working on shortening meetings, with the aim of resetting the cultural default for a meeting from 60 minutes to 30, and reducing the number of people who attend meetings.

Finally, the trial was designed partly as a test / demonstration of Microsoft Teams, their groupware product. Of course, there’s a long tradition in software companies of using your own products, as both a way to find bugs, and to prove the value of your product.

So what was the trial?

In its barest outlines, the trial was a month-long experiment to measure the impact of a four-day workweek in which 1) salaries were not cut, 2) the company made efforts to increase productivity or become more efficient, and 3) people were encouraged to do things outside work. (This being Japan, that kind of official sanction for extracurriculars does send an important signal to many people.)

August had five Fridays this month, so people have five extra days off. Technically these days were called paid leave, so salaries didn’t go down. The company offered subsidized weekend trips and technical courses, and would reimburse up to 100,000 yen spent on courses or for expenses incurred doing volunteer work. Just under 2300 people were involved.

The project was also meant to showcase the potential for Microsoft Teams to make work more efficient and people more productive.

So what happened? The company talked about a three-pronged approach to improving efficiency during the trial, and reported several measures:

Deleting: This category basically includes indicators consuming or spend less time, money, resources, etc. In this category, they report reductions in the number of working days (-25%), pages of paper printed (-57%), and electricity consumption (-23%). The last two are interesting partly because companies generally don’t try hard to measure these (and I confess I didn’t ask about them in interviews I’ve done). The electricity consumption number is especially valuable because our models of the impact of reduced working hours on energy consumption are based on relatively small data-sets so far, and lots of smaller companies that go to 4-day weeks don’t pay their own utility bill (it’s included in the rent), so the more data we have on this, the better.

Improving: This includes improvements in productivity, adoption of more efficient ways of working and collaborating. Labor productivity, measured by sales divided by the number of employees, went up 39.9%, but there were also increases in the adoption of 30-minute meetings (+46%) and remote conferencing (+21%). During the 4-day week, meetings were cut to 15 minutes, and the number of participants capped at five (though in practice the number of meetings increased slightly).

Satisfying: This is a measure of how people feel about the new schedule, and how their work-life balance changed. Overall 94% of employees reported positive experiences, and 92% were satisfied specifically with the 4-day week. (One partial exception was the sales department, which disliked being out of touch with customers on Fridays.) 55% of employees took summer vacations, and 6% took 2+ weeks off. There was a 3x increase in domestic travel subsidies for 1+ week vacations, and a 1.7x increase in sponsored activities.

Regarding Microsoft Teams usage during 4-day week, the highlights were replacement of email with chat (which theoretically is more efficient, though I’m skeptical of this, given the complaints I’ve heard of Slack overload); “time reform” to reduce waste; and support by Microsoft’s AI MyAnalytics, about which no one is saying very much.

It also sounds like Microsoft was working on helping employees build in time for concentrated work or “focus time” into the schedules.

Microsoft Japan says it will implement its own work project “Work Life Choice Challenge 2019 Winter” this winter, around theme ”Rest smartly, work in short time, enjoy the challenge.” Which could have been the subtitle of my next book.

To me, what’s significant here is that the experiment shows that a large company can implement (albeit temporarily) a 4-day workweek across a number of offices and functions; have substantial, measurable improvements; and create a prototype that they can continue to refine, and maybe make permanent, in the future.