Should you move into management?

Andrew Wulf’s My Biggest Regret As A Programmer was all over Hacker News a couple of weeks ago. In 1994 he could have gone into technical management or stayed a programmer, and, in his words, “I chose programmer because it was easier.” Twenty years later he thinks that was the wrong call, partly for money (his sister went into management early and, he says, has 10x his assets) and mostly because as a programmer he could see what was broken and had no power to fix it.

The comments mostly split into “yes, this” and “management is a different job and you’d have hated it”. I went the other way from him, around ten years ago. In 2006 I wrote a slightly smug post here about why managers get paid more than programmers, six months into managing my first small team, and reading it now is a bit like hearing a recording of my own voice. The conclusion was right, though: people are a much harder system to program.

What I gave up is easy to describe. I used to have days where I’d sit down at 9 and look up at 6 and something existed that hadn’t existed in the morning. That feeling is very rare now. My days are cut into thirty-minute pieces, the stuff I produce is mostly decisions and conversations, and you often don’t find out whether a decision was good for six months. For the first year or so I kept trying to get that feeling back by writing code at night, which mostly meant I was a tired manager and a part-time programmer who never had context. I still write code, but I’ve learned to keep it off the critical path. If my PR is blocking a release, something has gone wrong.

What I got is what Wulf is describing from the outside. When something is broken, like the deploy process or the way we interview or a team that’s been quietly miserable for months, I can actually change it. That part lived up to what I expected. A good hire is worth more than anything I could write myself in a year, and a bad one costs more than anything I could break. I also learned a lot about myself, not all of it flattering.*

So, should you? I’d skip the question of which one is better and ask a couple of narrower ones. Do you find yourself caring about why the team is slow, as opposed to just being annoyed by it? When someone on your team is struggling, is your instinct to fix their code, or to figure out why they’re stuck? If it’s the second one, you’d probably enjoy at least parts of the job. If you just want the authority and the salary, you’ll get them, and you’ll be bad at it for a while, and the people reporting to you will pay for it.

The other thing I’d say is that it’s more reversible than people think, at least for the first couple of years. Going back to an individual contributor role after a year of managing isn’t failure, and you come back a better engineer, because you finally understand why the meetings exist. After five or six years it gets harder. Your skills drift, and the market starts seeing you as a manager whether you like it or not.

I don’t regret it. Some Tuesdays I’d trade it all for a week alone with a debugger, and I’m told that’s normal.

* Mainly that I avoid conflict a lot more than I thought I did, and that “I’ll just do it myself” is a habit you have to break on purpose.