Startups And Work/Life Balance

For the last seven years, I have lived and worked in Japan.  I love many things about living here, but the working culture is toxic in at least one fashion: the societal expectation for professionals is that one spends substantially all of one’s time at work.  This is accomplished largely by subtle social control.  Our employment contracts say things like “Work hours are from 9 to 6.”  The government has an office in charge of reducing amount of hours worked.  There are posters exhorting employees to not overdo it. Nobody walks around and bars the door at 6.  And yet, no one leaves, because leaving at 6 is not done.  It would communicate that one was not sufficiently invested in the work, one would be abandoning one’s teammates in their hour of need (regardless of whether they, too, are merely marking time until 1 AM), etc.

 

I quietly tolerated this for years, because it is the Way Things Are Done here, but now that I’m my own boss I can say what I really think: this is evil.  If it were just a wasteful tradition passed down from time immemorial, it would merely be stupid.  But it negatively impacts the lives of employees and their families and loved ones.

 

Startups Are Almost As Bad

There is a nearly pervasive attitude in the startup community that work should consume one’s life.  Jean Hsu recounted an anecdote on her blog recently http://www.jeanhsu.com/?p=296 where someone suggested that successful companies, by necessity, have employees work through dinner.  This is widely viewed as the proper way to do a startup, and if you work on a different schedule, you are likely to be told you are “doing it wrong” by employers, investors, and peers.  This is peer pressure deployed as a weapon. 

Partially it is simple hazing: the experience of shared suffering in the group context is believed by many to form stronger bonds.  (And, as with classical hazing, “I put in my dues, now you have to put in yours or I will look pretty freaking stupid now won’t I” causes the system to perpetuate itself.)

In addition to the hazing rationale, some portion of this attitude is just pure exploitation.  Just like soulless megacorps like i-banks and Google give out free dinners to encourage people to stay around for them, establishing a norm that salaried employees spend the majority of their day at work theoretically allows companies to squeeze extra productivity out of a fixed amount of wages/benefits.  [Apologies if the soulless megacorp you work at or admire has a good PR team.  In that case, pretend I am talking about some other soulless megacorp because yours is clearly a lovely one-off exception, and send the PR team an email saying “Mission Accomplished.”]
 

Punishing Hours Do Not Improve Productivity

Paul Graham describes a law of conservation of work (http://www.paulgraham.com/wealth.html): if you want to make a million dollars, you can either work for the Post Office for 50 years or compress all that work into 3 years.  This has much to recommend it as allegory: it is punchy and evocative as a metaphor.  However, factors like the amount of leverage one has in their work and savvy in negotiating matter much, much more in determining how much you make than the amount of aggregate suffering you experience.  (For example, many engineers could earn a sizable amount of money just by remembering to make a counter-offer.  That’s all.  One sentence.  The job is the same, the candidate is the same, and the salary meaningfully increases just because you asked for it.  For another example, no engineer in Silicon Valley suffers anywhere near as much as the folks who make your iPhones or clean your toilets, and you will make many, many times what they do.)

We are finally starting to measure and iterate in startups, and this is producing huge amounts of wealth.  We are fast approaching the point where doing e.g. conversion optimization based on whimsy rather than evidence is laughable, because of the clear, reproducible, undeniable benefits of e.g. A/B testing.  It provably makes big stacks of money.  We have not yet started trying to measure productivity in the same way – the seminal studies like Mythical Man Month date back to the 1970s.  Many things about our working culture smack of mysticism and witchcraft: the notion that an industry as diverse as startups shares the same time/competitive constraints, the idea that all job roles require approximately equivalent schedules (“every free hour”), the relative static-ness of schedules without regard to the product life-cycle, the “emergency”/”crunch” periods which seem to be triggered every single week, etc, all point to management by tradition as opposed to trying to study and understand the problem.