This post has been sitting in my head for a while. But recent news about another 4-day workweek study reminded me that now is the tell to share my tale.
A few months ago, I concluded that a 4-day sprint or a 4-day work week is better for us. And it's not just because this means we can produce less work, but rather, it is because it allows us to produce better work. In the study linked above, revenue increased. But how can that be? Every week as a software developer I feel pressed for time, unable to finish everything that I promised I'd be able to do. Clearly, removing another day of work will cause me to be even further behind than I already am?
Lucky for you, I've been able to run an experiment these past few years and gained some insight into what's going on.
Background
About two years ago, I decided that I was unlikely to get a significant raise at my job. I thought it would become more likely that I could argue to work only 4 days a week instead of 5. To "prove" that such an arrangement would be feasible, I started to slack off one day a week, putting in fewer hours and less effort. I then asked my manager if they had noticed that I was getting less work done during those weeks. After agreeing that I was not getting less work done, I was able to negotiate a few weeks a month at only 4 days a week. Then I switched jobs and joined a company that had a company-wide policy of having 2 short weeks a month. I have since left that job and gone back to working a 5-day work week. This means that I have been able to compare what it's like to work shorter weeks vs longer weeks in both short iterations and long iterations.
Theory
I thought about why is it, that a 4-day sprint or a 4-day work week is doable? Are we wasting that much time during the day? Does a 4-day sprint cause me to spend more hours during those 4 days? Perhaps this is just a Hawthorne effect? Why did I feel more stress during a 5-day week than a 4-day week? Let's take a step back and think about the development process.
As is commonly said in the industry, the hard part of software development is in the thinking, and the communicating, and not the typing. The syntax is now easy to look up. Communicating and thinking, solving problems, and finding the right structure or the right algorithm is the hard part of our work. When work goes poorly, most of our day is spent reading code and trying to understand why it's not doing what we expect. Sometimes, we spend hours and hours of trial and error, thinking only a short amount of time has passed. Other times, we are so into the groove of things, hours go by and our "quick solution" appears to work perfectly. Other times, we spend our days waiting for a Code review or some other bureaucratic process to continue so we can get back to work. But always, we are 'busy', we are working hard, and feeling that we aren't working fast enough. Sometimes we even get to a situation where we try to 'hurry up and wait.' What this means for our thinking brain, is that we are constantly putting our brains under stress. Thinking and problem solving or resting and worrying. As a result, we don't have time to properly allow thoughts to percolate through our system.
One possible solution to this problem, and sometimes it works well for me, is to practice Pomodoro techniques. Set a time, every 25 minutes, take a 5-minute break. Occasionally break for 10-15 minutes, get lunch, etc. In theory, this should allow us to work 6 days a week, with enough time to allow ideas to percolate. We can get "shower thoughts" throughout the day by taking proper breaks. From my experience this works well... but not for 6 days a week.
Problems with Pomodoro
There are three main problems with the Pomodoro technique.
We have a habit of pushing off the breaks when we are in the middle of something juicy. When you think you are in the flow, or you think you are close to being done, you are very likely to push off the break. (You are also very likely to be further from your goal than you think.)
Your breaks get interrupted by other issues or work. Perhaps you notice chat messages or new emails. You get pulled into social media or otherwise engage your brain during that break.
15-25 minutes is not a long enough time to really rest. When you get your "shower thought" or "commuting thought", it might be during a 15-60 minutes window, but it's actually during a much longer winding down period. You've spent hours sleeping, or an hour or so of being in a routine before those thoughts come to you. You need more rest than Pomodoro will allow.
Pomodoro is great, but it's not sustainable 6 days a week, and as it turns out, not 5 days a week either.
No Rest for the Wicked
We spend 33% of our week sleeping. We probably also spend a minimum of 25% of our week working. That leaves another 42% for us to spend time alone, with family, with friends, adulting, and for side projects. If we divide that up evenly, that gives us less than 2 hours a day for each of these non-work related activities. I don't know about you, but for me, 2 hours is barely enough time. The result is that during working hours, my mind is occupied with other thoughts. Did I pay my bills? What will I do tonight? Is there an appointment I forgot to block in my calendar? These worries and concerns, take up precious thinking time; time that could be used to subconsciously solve my trickiest software development problems.
Moving to a 4-day work week means the equation shifts. Now we spend 33% of our time sleeping (hopefully) and less than 20% of our time working. This gives us almost 50% of our time to devote to the rest of our lives. In addition, we now have 2 extra 8-hour blocks to devote to adulting and hobbies and friends/family. I don't know about you, but for me, I can get a lot done in 8 hours. This frees up my mind, gives me less to worry about and lets me rest. In the end, this means your brain has plenty of time to solve problems subconsciously.
For myself, instead of "shower thoughts", I often came back to work each week with a short list of "weekend thoughts". When I work 5 days a week, I only get these "weekend thoughts" during times of crisis.
If 4 is good, 3 or 2 must be better!
I can't say I've ever had the chance to experiment with a 3-day work week, but I was able to experiment with a 2-day work week. Weekends are great to get some things done, but from my experience, you can't get enough done to get good feedback promptly. Each weekend, you come back to work and have to relearn what you did last time. Those 2 days are very effective, and a lot gets done, but not enough to make up for the other 2 days you aren't working. It might be worth experimenting with it though. Perhaps 3 days is enough time to deal with all the back-and-forth communications required to get an iteration through to completion.
TL;DR
A 4-day sprint gives us the following benefits:
Our minds don't worry about errands, adulting, family, or friends. We can focus on work and not try to pay our bills during working hours.
Our minds do heavy thinking during our downtime. Each week, we have time to let our brains process difficult problems, or come up with innovative ideas.
4 days is enough time to get a sprint's worth of work done. Because our work is more effecient.
3 days may not be enough. 2 days most certainly isn't.
A 5-day sprint causes the following problems:
Too much time spent worrying about non-work related responsibilities.
Not enough time to let our brains sit and think and digest.
Being stressed and tired, causes our work to suffer in quality.