Joining a new company
I decided to publish an email I was writing for a buddy - he's joining a company for the first time (though he has successfully run one). Given I've joined about 10 companies now I've picked up some some key strategy and decided to share them with him.
---------- Forwarded message ----------
From: Michael Ossareh <ossareh@gmail.com>
Date: Sun, Apr 1, 2012 at 08:31
From: Michael Ossareh <ossareh@gmail.com>
Date: Sun, Apr 1, 2012 at 08:31
Be non-political
This sound easy - but it is actually hard. We're leaders, that is why we start companies. We're charismatic, that is why we could raise money. Leadership and charisma cannot be taught, they're acquired when we're young - usually around 5 to 6 years of age. When you join a company being a leader is dangerous because there is already an established hierarchy in the company; politics start taking place once you're about 5 people, and the increase in politics is usually larger than linear as you hire more people.
Be moral
I've found that when negotiating politics sticking to the right thing is the best. Good companies have less of this issue because they want to do the right thing however no company is full of A+ people, this is just fact. The non A+ people will hopefully want to be A+ people and fight for what is right, but sometimes they don't and they make life hard. Always stick to what is right.
Take direction
Knowing what is the right thing is hard. Analyze people to establish whether they do the right thing, I usually do this through a process of working out where other people are in terms of my thinking. This sometimes makes people think I'm dumb, because I ask simple questions, but their answers permit me to plot them on a graph of moral correctness. When I work out who I think is right in terms of my definition, I then take their direction with regard decisions made within the company. This helps you fall into step with the company and as a result gain institutional knowledge faster, permitting you to make the correct decisions about why, when and how to do something.
Be independent
When faced with a problem, decide a reasonable length of time to work on it before you ask for help. Assume you're asked to do "foo" and as you start with foo you realize that you have a problem, perhaps tests are failing or maybe there is some gnarly code and you want to fix it but it is touched by a lot of other things, decide how long you're willing to work on it without help; I usually set 3 hours for hard problems. Be strict with that box, once you hit the limit then ask, but in that time box work hard on it. The benefit to this is that you acquire a lot of knowledge from the edges of the problem, then when you're given help you can actually help the helper.
Establish expectations with your partner
You're moving from giving her lots of time, to giving her less time. Without fail every single time I've started at a company I've heard complaints around "I don't see you anymore", "you're grumpier since you've been working at X", etc. This is because I throw myself into things and want to impress people. It is like starting a company, except instead of building a user base you're building a trust base. You must build this in a relatively short length of time; in the order of 3 months or so. If you do not build that trust base, your desires to have your opinion heard (remember, we're leaders) will not be satiated and that manifests itself as frustration. Your partner is going to be most vulnerable in this situation; in one case my partner went on to cheat on me because she lacked attention - she didn't even know she was doing it. So set expectations, let her know you'll be busy, when to expect you home (and stick to it - or let her know if you cannot), organize date nights (and stick to them).
Choose a cause and champion it
Never have I believed it was correct to hire someone with a complete intersection of my own knowledge. Knowledge is a pool and you want it to be as wide and deep as possible; each employee brings with them new potential. Hiring is actually quite beautiful; your company becomes a patchwork of brains. You'll want to satiate your desires to be heard (we're leaders!) and one of the best ways to do that is take something the company isn't good at that you are either good at or are willing to become good at (i.e. you've an interest in it) and champion it. When you're putting in all those extra hours, most of them should be spent on this. I arrived at JTV and it was coding-cowboy-corral (no tests, no code review), I knuckled down and built testing frameworks, I wrote emails about how to test, why to test, I pair programmed tests with people, when helping people to on board I explained the importance of testing and code review. 12 months later I see people understand why, and they contribute their own tests and improvements to the testing infrastructure and we have zero deployments to production without at least one other set of eyes sanity checking the code; it is really satisfying. I'm moving my maniacal focus to tooling now :)
I hope this helps,
Cheers,
mike
