Since it's known from books like mythical man month from over 30 years ago that more hours very rarely translate to better code or faster delivery, why do I keep reading about companies like FB and quant firms doing it here? Isnt the managers well educated in the area or was developers before? Why don't you show and tell them how it is? My brain works most time anyway but office time above 40h +-10% is pointless
Impact matters. Not hours. You may have to put in more hours in the process of impact.
For the same reason why they think savings the planet = self driving cars and not allow their workers to work remotely, thus saving the commute.
Many of the books written by people who didn’t build companies like Microsoft, Tesla or Amazon or probably didn’t work in important positions in similar companies. Even Marissa Mayer had to pull in all nighters at Google! I’d be interested in a book authored by Elon Musk on this.
Tmmm is written by a guy who wrote one of the first successful mainframe systems.... But I'm sure Elon "funding secure" musk has more knowledge on the subject
Because companies need more than appeals to authority from a 30 year old book before they fundamentally change business models. And, most startup founders with companies larger and more profitable than IBM aren’t going to listen to some IBM lifer about how to run a successful company.
Tell me again how Uber Tesla or snap at profitable
You mentioned Facebook. Now you shift to other companies that aren’t losing money due to overworked employees and argue those employees should work even less? I’m not following where you are going.
They don't care about long term retention unless you can keep that pace up. So if their average is high because they're burning people out, why would they care? It takes a massive change in hiring practices and retention policies to change that. I'm pretty sure the first wave of millennials is feeling the burnout hard already, and I suspect that as the decade progresses we will continue to see paths towards better working hours simply because all of those good people are already migrating to companies with better policies in place for it.
Yes but it doesn't even increase profits, which is the end game in theory
They don't. The notion of more hours = more output is plain and simple propaganda to hide a simple truth: it is cheaper for them to extract a few bits more of work from the same employee instead of hiring a new one and splitting up the work to be done. Only when things don't scale anymore they start looking for new hires.
Ok that's a good point. As a means to avoid hiring
Statistics are about populations, not individuals. There are people who only have the motivation and mental stamina to work 20 hours per week without burning out. Others can maintain 60 per week. To determine that all humans have the same cognitive limits is as absurd as claiming that all track athletes can run only the same distance.
yes, there can be outliers. But most software are made in teams, and those like I have said are measured. Then you have the overhead of documentation, bringing in new developers, complex systems that break down more often because they are bigger and so on And with your running example, it's only when they run it matters. No one can run say 50 km per day in marathon speed every day, which is more what I refer to. But let's go down to stats instead, is there any scientific basis for more hours = better products or software?
Running 50km per day is typical for competitive ultramarathoners. In terms of percentiles, it would probably be equivalent to at least 90+ hours/week for software engineers. It is sort of an extreme example. On the other hand, I don't see any scientific basis for 40 hours ± 10% being the optimal for all software engineering, which is part of the original point raised that carries burden of proof. I have personally worked 60-70 hours/week productively, although I had to make other sacrifices in life.