Almost every pitch for a four-day week includes the same sentence: “Microsoft Japan tried it and productivity rose 40%.”
It is a striking number. It is also the most misquoted statistic in working-time debate. Not invented, but detached from its context so thoroughly that it now promises something no serious pilot should promise.
At Day5Group, we help organisations transition to a four-day week, and we deliberately talk clients out of quoting it without qualification. Credibility is a consultancy’s stock in trade; this number, used carelessly, spends it.
What actually happened in August 2019
Microsoft Japan ran a Work-Life Choice Challenge for one month, August 2019. Its roughly 2,300 employees were given Fridays off, with the office closed, as a special paid arrangement rather than a permanent contract change. The month came with supporting measures: meetings were capped at 30 minutes and five attendees in guidance widely reported at the time, remote participation was encouraged, and staff were urged to use the time for rest, learning and family.
Microsoft Japan then compared the month with August 2018 and reported that sales per employee had risen by about 40% year on year. It also reported electricity use down about 23%, printed pages down about 59%, and that 92% of employees responded positively. Staff also took less other time off during the month.
Those operational figures are real company-reported results, and the meeting discipline alone is worth copying.
What the 40% does not mean
It does not mean individual output rose 40% because people worked four days. Sales per employee in a single summer month is a volatile measure, shaped by the sales cycle, product timing and how revenue lands, as much as by effort in those four weeks.
It does not mean a permanent 40% gain. The challenge lasted one month, in August, a holiday-season month in Japan, partly alongside existing leave patterns. Novelty and goodwill are powerful over four weeks.
And Microsoft itself has since urged caution in how the figure is used, noting the sales result should not be attributed to the challenge alone. That qualification rarely survives into the pitch deck.
Careful readers will also notice the base. A reported rise in sales per employee is not the same as measured productivity per hour across a year. No controlled comparison sat behind it.
Why the number still matters
Discard the slogan and the case remains instructive. A large, hierarchical, client-facing business closed its offices on Fridays for a month without the sky falling. It paired the closure with meeting limits and reported less waste: less printing, less electricity, fewer sprawling meetings. Employees approved, overwhelmingly.
That is a useful, defensible claim. The indefensible one is to take a one-month, company-reported sales comparison and present it as the expected return on a permanent scheme.
Compare it with the evidence we treat as stronger. Perpetual Guardian’s 2018 trial was academically monitored and showed stress down, work-life balance up and output maintained. Iceland’s multi-year public-sector trials showed service maintained or improved at scale while health measures improved. Neither case offers a magic percentage. Both offer something better: a pattern you can test in your own organisation.
How to use the statistic honestly
If you must cite it, cite it whole: “Microsoft Japan reported sales per employee up about 40% versus the same month a year earlier during a one-month August 2019 challenge, and later cautioned against attributing that sales figure to the scheme alone.” Then stop. Do not forecast your own 40% from it.
Instead, make the claim the evidence supports: in well-designed trials, output has generally been maintained while wellbeing improves, and firms fund the time by cutting low-value work. That sentence will survive due diligence by your finance director. The 40% slogan will not.
Then design your pilot to produce your own numbers: baseline, output measures, a customer floor, actual hours worked, and a decision date. One honest internal result is worth more than a borrowed headline.
Sources
- Microsoft Japan Work-Life Choice Challenge, August 2019 — reported by NPR, 4 November 2019: https://www.whro.org/news/2019-11-04/microsoft-japan-says-4-day-workweek-boosted-workers-productivity-by-40
- Perpetual Guardian trial, New Zealand (2018), academic monitoring by Auckland universities — reported in ScienceAlert, 2018.
- Iceland public-sector shorter-hours trials 2015-2019 — Autonomy and Alda, reported July 2021.
Day5Group is a consultancy that assists in the transition to a 4 day work week. More about Day5Group · All blog articles