People talk a lot about “the startup life” and how it is to dive into it from regular employment, but what about the other way round? Having recently completed a rare stint managing developers at a large-ish company of about 400 people (don’t laugh – that looks big to me) I wanted to share a few thoughts on how it is to go from raw startup life, to life in a big company.
The best part of managing devs at a big company is that if you do it right you can have a substantial number of people writing production quality code together at the same time, doing great big things you can’t dream of at a two-man startup. I never fully appreciated the joy of that phenomenon until I went big. No demos, no wild pivots, no fire drills, no due diligence, no clueless angels or arrogant VCs, actual QA, a Build Guy (Hi Mike!), no working with crazy people and C players because they’re the only people crazy or desperate enough to work with you. Just big monitors, pricey tools and people who know how to use them going full bore. The company was hugely successful, the group was hugely productive and for the most part, people enjoyed working there. I enjoyed the hell out of that part.
Still, it had it’s downside.
Lesson #1: You are irrelevant Part I. There is nothing you can do all by yourself. Even if there is, you won’t be allowed to do it all by yourself so … there is nothing you can do all by yourself. I often found myself reveling in (i.e. milking) the tiniest little tasks simply because I could do them by myself on my own authority. The depressive effect of this lesson is offset by …
Lesson #2: You are irrelevant Part II: you are actually irrelevant. Unless you are prone to huge mistakes, your presence or absence will not affect the business in any material way. This can be comforting if you have a long commute, enjoy long lunches or if you like to turn your phone off when you go on vacation. But in those fleeting moments between arriving late and going to an early lunch, if you feel like, you know, doing something, you can apply …
Lesson #3: The most productive thing you can do is to hire and keep productive people, and fire unproductive people. If you’re in a one-new-hire-every-couple-of-months growth mode, the hiring takes maybe ten percent of your time. The firing, maybe a couple of days per year. This still leaves you lots of time for applying …
Lesson #4: Keep your crew happy. In some organizations, happy people are not productive. For example, prison guards are happier when they’re not watching prisoners. Fortunately, software development is an addictive activity and thus its practitioners are happiest when working. The faster they are working, the happier they are. The harder they are working, the happier they are. Whether you want monstrous edifices of code, enormous slinky snakes of elegant functionality, or lots of little hacks, well … that’s all the devs want … to be allowed to make these things for you and your only job is to let them do it. On paper, it’s easier than being vice president. Dan Quayle could do this job. Mostly it requires getting the developers to tell you what needs to be done, telling them to do it, and then resisting the urge to tell them how to do it**. In general, this makes them happy. As Dilbert said when HR installed a lab-rat feed-slot in his cube wall:
This job is so stressful … but the pellets are excellent
You can hang out with the crew, or brood in your office but all this management by walking around bullshit doesn’t matter. The only time the crew really needs to see you is when you’re applying ….
Lesson #5: Make enemies of anyone who disrupts “the flow”. The meeting-callers for example. Being just a manager who by definition had nothing better to do I used to go to meetings in place of my guys whenever I found them invited to some bullshit meeting. This cut way down on the meetings, bullshit and otherwise, especially when I told the other participants that I came in Sally’s place because Sally had better things to do and I didn’t. I made it clear to my guys that I would forgive them for writing bugs, but not for wasting time in meetings. It got to the point where my guys apologized to me for going to meetings, or worse for calling them. This is the way you want it, but it will make plenty of enemies.
Then there’s the team-organizers, the guys who want to form a flying squad for every little project that comes into their heads. This squad deals with a single issue and the organizer feels like a hero, then the members go back to what they were doing before … assuming they can remember what it was and why they were doing it and that whatever it was didn’t fall apart because they couldn’t keep a team together long enough to finish it.
Watching the flying-squad phenomenon in action reminded me of those corny TV shows where everyone gets together to put on a show to raise money to put a new roof on the orphanage. Let’s have a show!
Well actually, let’s not. When you see one of these trains coming down the track, your dignity as well as your membership in the Agile community requires you to throw yourself in front of it. Fortunately, as a manager you have nothing better to do than lay on the tracks, waitin’ on the Double E, and this on-the-fly team would probably have lots of meetings so squashing it is a two-fer. Gumming up the wheels of the locomotive and bleeding all over the tracks ensures that you won’t go far in the company, but even if the train runs you over, you will certainly go to Agile heaven. And going far in the company isn’t an issue anyway because of …
Lesson #6: So after a) hiring a bunch of good people b) realizing that you’re irrelevant c) letting the devs do pretty much what they think needs doing and d) making enemies of everyone in the company who likes to interrupt the flow, the smartest thing you can do is to not overstay your welcome. You’re a startup guy. Every big company needs a dose of that now and again, but too much is toxic. Don’t be toxic.
** Okay, there IS also the small matter of getting Product Management to organize their thoughts, but if you can’t do that you probably shouldn’t even be vice president, let alone a dev manager.