The archive · Work & Ways of Doing · Operational decision · 2007–2008
37signals cuts to a 32-hour week and finds four days do the work of five
In 2007-08 37signals cut Fridays and ran a 32-hour week, finding that 'just about the same amount of work gets done in four days vs. five days.'
37signals (Basecamp)
What it had to solve
37signals, the small Chicago software company behind Basecamp, wanted to be one of the best places in the world to work, learn and be happy. A summer experiment in 2007 suggested the team produced about the same output in four days as in five, so founder Jason Fried made the shorter week a standard for 2008.
How it works
In December 2007, 37signals — the Chicago software company behind Basecamp — decided 2008 would be a year of workplace experiments. The previous summer the team had quietly tried four-day weeks and found that 'just about the same amount of work gets done in four days vs. five days.' In March 2008, founder Jason Fried announced the four-day week as the company standard: Fridays off, with customer support and emergencies still covered.
The idea was not to cram 40 hours into four days. When Forbes ran a piece titled 'Why A Four-Day Work Week Doesn't Work' — built on four ten-hour days — Fried pushed back: the point was a 32-hour week of normal-length days. 'Results, not hours, are what matter,' he wrote, and 'once everyone has less time to get their stuff done, they respect that time even more. People become stingy with their time and that's a good thing.'
The shorter week sat alongside other experiments — 37signals helped pay for employees' passions like flight lessons and gave everyone a discretionary credit card — all designed to treat people like adults. What made the four-day week the lasting idea was that it reframed working time as a constraint that improves the work itself, not a benefit that reduces it.
Why it lands
- Dropping a fifth day forces teams to question which meetings, interruptions and tasks actually matter, so output holds while hours fall.
- A standard Friday off creates predictable overlap on working days, so collaboration doesn't quietly disappear.
- Running the change as a public experiment with results invited debate and copying, turning one small company's schedule into a movement.
- Keeping support and emergency coverage on Fridays showed the change was about focus, not about vanishing when customers need you.
What it did
The company found output held steady, and the experiment became a public reference point: Fried defended the 32-hour week when Forbes attacked the four-day idea, and the posts drew hundreds of comments from teams that copied the practice.
What you can take
Hours are a proxy for output, not output itself — cut the time budget and people cut meetings, interruptions and busywork, so a shorter week can be a forcing function for focus rather than a perk.
Since then
The four-day week became a permanent part of 37signals' identity, and the two 2008 posts are still treated as the canonical first-hand account of the experiment. As the company grew and later renamed itself Basecamp, it kept citing the shorter week as proof that 'results, not hours' can work — and the experiment became the standard reference point for later four-day-week trials, including Microsoft Japan's 2019 test and Iceland's national pilot.
Sources
spotted an error? The archive wants to know.
Your turn
You just read one. Describe the brief you are staring at, and see who has been given the same problem.
Free account · 3 free questions · no card