"Keeping the lights on" work needs to be refrained as investment work.
It can be tricky to ascertain which investments are worth making -- THIS IS WHAT SENIOR ENGINEERS ARE FOR -- but good investments pay off a thousandfold or more.
It can be tricky to ascertain which investments are worth making -- THIS IS WHAT SENIOR ENGINEERS ARE FOR -- but good investments pay off a thousandfold or more.
If you are an eng manager, it's your job to work with senior eng to figure out what investments to make. Then provide air cover.
And advertise the hell out of your successes, so that you earn more currency/credibility for next time.
And advertise the hell out of your successes, so that you earn more currency/credibility for next time.
This last point seems to get lost a lot.
If you did a hard rewrite that's still paying dividends months or years later? KEEP TALKING ABOUT IT. Keep rewarding the people who pushed it through against all odds. Praise the actions you want to see more of.
If you did a hard rewrite that's still paying dividends months or years later? KEEP TALKING ABOUT IT. Keep rewarding the people who pushed it through against all odds. Praise the actions you want to see more of.
The problem with infra work and investment work is that when it's done well, it tends to be invisible.
Your job as a manager is to lean hard against tendency. MAKE it visible. Point it out, celebrate it, repeat til long after you're sick of saying it.
Your job as a manager is to lean hard against tendency. MAKE it visible. Point it out, celebrate it, repeat til long after you're sick of saying it.
And when you made bad investments? Don't hide those either! Post mortem them, let the org see you learning from your mistakes. Bring the team along so everyone levels up.
This builds trust in your team, too. People need to see an accounting for the currency they give you.
This builds trust in your team, too. People need to see an accounting for the currency they give you.
And if they're letting you page them at 3 am, they need to trust you... a lot. You need to prove yourselves constantly. Prove that you value their time and attention and social life and sleep. Prove that you learn from your mistakes. Prove you can listen.
The way to combat short sightedness is by extending your horizons. Senior engineers and managers are the institutional memory of the team. Anyone with empathy can be the conscience of the team.
You need to exercise both. And reward them right up there with feature work.
You need to exercise both. And reward them right up there with feature work.
Because ultimately, all these investments are what will actually let you keep shipping feature work swiftly and safely.
Without them you will grind to a halt for opaque and frustrating reasons at the worst time. Without them you will slow down and drive good people to quit.
Without them you will grind to a halt for opaque and frustrating reasons at the worst time. Without them you will slow down and drive good people to quit.
All of you who are tweeting at me all the elaborate reasons why you can't be on call bc it sucks so much: I hear you.
On call is not a panacea for all dysfunction. It's just a really powerful tool, to be creatively deployed as part of a full suite of solutions.
On call is not a panacea for all dysfunction. It's just a really powerful tool, to be creatively deployed as part of a full suite of solutions.
I tend to jump straight to the inflammatory conclusion -- "test in prod! put all devs on call!" -- because it forces the conversation and gets the ball rolling.
But I'm not crazy. It's a lot of work to get there. No shame to those who are starting many steps back.
But I'm not crazy. It's a lot of work to get there. No shame to those who are starting many steps back.
