Skip to main content

Incentives

Every kid has, while grumbling about some injustice perpetrated upon them by their parents, declared “when I’m a parent, I won’t make my kids do these lame things. I’ll be a cool parent!” Sometimes they’re right. Most of the time, however, they are wrong. Why are they wrong? Because they don’t understand that as a kid, they have fundamentally different incentives than as a parent. A kid may want to eat lots of candy and not go to school, because their incentives are to get dopamine out of eating sugar and playing. As a parent the incentives are to ensure that their kid is healthy (i.e. not eating too much sugar) and educated (i.e. goes to school). These incentives are fundamentally at odds.

The same is true for programmers. Many a programmer has thought at some point or another, “when I run a company, I won’t make programmers do these lame things. I’ll be a cool boss!” These lame things, instead of going to bed early and eating healthy, is stuff like setting deadlines, writing imperfect code in order to ship, letting customers determine the roadmap, and so on. In a programmer’s mind, they’ll run a company where the code is done when it’s done, plenty of time will be devoted to refactoring, and no pesky PMs will tell them what to do.

You see, the incentives of a programmer are to write code in a way that feels good, and to have the freedom to write what they want. The first point is worth some more explaining. At least for me, there is certain code that feels pleasurable to write and other code that feels less pleasurable. Refactoring, for instance, is a very pleasurable activity. This is not surprising. Refactoring is fundamentally an attempt to “clean up code” and therefore make it more understandable. Of course, this really means more understandable in the eyes of the person refactoring. Combined with the general aura of virtue that surrounds refactoring, means that programmers are very much incentivized to refactor. They get to remake the code to be easier for them to understand and they get to feel superior for doing so.

This focus on pleasure can also apply to languages. I like writing Rust. It’s a fantastic combination of a language that lets me express my ideas, that challenges me but doesn’t frustrate me, and that lets me write fast code. If I were starting a company, I’d be very tempted to write my code in Rust.

Should I though? In most cases, probably not. If I’m writing a back-end server, I’d probably spend more time getting my server library connected to my auth library to my database than I should. Something like JavaScript or Ruby on Rails would be a lot more pragmatic, at the cost of being less “fun”.

And really, the latter point of freedom is just an extension of the former point; in order to spend time doing the pleasurable tasks, one must have the freedom to avoid the less pleasurable ones.

The incentives of a company leader are quite different. The company leader needs to build a product that people actually want. The leader needs to promise stuff in advance, then hit those deadlines. And if the code is a little messy or there’s a small bug or two? Not a big deal.

Neither of these people are in the wrong or right. Both of their incentives are valid. A programmer needs to fight for quality because a leader will likely not. A leader needs to fight for customer needs because a programmer will likely not. It’s important that we have this balance. And indeed, both parties have an incentive to keep the other party happy, and therefore compromise.

Which does mean that when this balance is disrupted, say, when a programmer becomes a company leader, there is a real problem. Unlike a parent, a newly minted leader doesn’t always understand this change in incentives. They may continue to play by their old programmer incentives. What’s especially dangerous here is that they won’t receive any negative feedback for these actions. The other programmers will cheer them on for providing what the programmers want. Only when reality comes knocking in the form of no users or no funding will the new leader realize how much their incentives have shifted. Even then, there is some danger that the leader still won’t change because they’re afraid of looking like they “sold out” and became an evil manager.

Which, again, managers are not inherently evil. Sure, there are managers whose methods hurt the company. But there are also programmers who hurt the company: programmers who spend endless time refactoring and not shipping; programmers who refuse to talk to users out of hubris; programmers who pick tech stacks out of their personal interest and not out of pragmatism.

Going back to the parent analogy, you can have bad parents and still recognize the utility of parenting.

Therefore, if you’re switching positions, whether it’s becoming an engineering manager, starting a company, working as a PM, etc., you should probably reassess your incentives and act accordingly. And begrudge people their incentives either! Just because they’re not doing what you want, doesn’t mean they’re bad at their job or out to get you.