3 ms·

> When you make a decision as a manager or a director, the impact of that decision has a multiplicative force. Every time I hear this explanation, I can't hel

> When you make a decision as a manager or a director, the impact of that decision has a multiplicative force. Every time I hear this explanation, I can't help but think "so we're still smoking Luck Strikes, huh?" When you make a decision as a software engineer, the impact of that decision can often have a multiplicative or even exponential force. The goal of 21st century organizations -- especially software organizations -- should be to make things scale in way that's at most sublinear with the number of people you throw at it. Automation prints money and businesses that can harness automation have huge profit margins. People are expensive and body shops are shitty companies waiting to be replaced when technology catches up with whatever work they're throwing bodies at. It's not 1960 anymore. Invest in your automators and treat your people managers as a low-value-add cost center.

TameAntelope 6y ago
One question for you: how does an engineer know what to automate?
- throwawaygh 6y ago
  How does a people manager know what to automate? And, more importantly, what might be the massive unintended consequences of routing your org's permission structure in the same way you route its head-count? Strategic decision-making is not the same as people managing, and people managing can even create blind-spots that get in the way of good decision-making.
  - TameAntelope 6y ago
    A manager (who said anything about people manager? I even explicitly said otherwise) knows what to build because she's coordinated with the groups who spend their entire day focusing on the customer, as well as the other developer groups to make sure everyone isn't working on the same thing, as well as leadership so they know what's being built and how it aligns with their goals and vision into the market (because that's what their expertise is in), as well as within her own team to make sure people are building things that can work together smoothly/cleanly, as well as covering for people who have illness/unavailability, and so on... Would you rather have developers wearing every hat in an org? When would they find time to code? I know many developers who don't want to have to do all of that work simultaneously, and are extremely grateful for a manager who can take many of these tasks off their plate. When was the last time you tried to add a recurring meeting to a development team's calendar? They... don't like it, in my experience, and I respect the hell out of that.
    - throwawaygh 6y ago
      Where did I say we don't need managers? I explicitly said otherwise: Invest in your automators and treat your people managers as a low-value-add cost center. Just because you treat something as low-value-add or admit it's a cost center doesn't mean you don't need that work to happen! > When was the last time you tried to add a recurring meeting to a development team's calendar? Yes, I agree, managers' secretarial and political work is incredibly important to the success of an org. But the 1960s idea that it's a "force multiplier" worth high comp is a mistake. I'm not contesting whether managerial work is important or necessary or to be respected. I'm contesting the idea that managerial work is a force multiplier for individual contributors, in any sense other than in the same sense that janitorial work or admin assistant work is a force multiplier. The only orgs where I've seen managers to be actual force multipliers were body shops.
      - TameAntelope 6y ago
        A good manager is worth 10x a good developer, and a good developer managed by a good manager is 100x base developer value. This is why managers often get paid more, and it applies even more the further up the chain you go. A good CEO is worth every single penny of the millions they get paid, because of the multiplicative force they have on everyone below them on the org chart.
        throwawaygh 6y ago
        > This is why managers often get paid more I think we are disagreeing because we're talking past one another. Hopefully the following observation helps make my point: The average base pay for a manger in the USA is much lower than the average base pay for an engineer in the USA. I'm talking about management as a discipline in general and engineering as a discipline in general. > A good manager is worth 10x a good developer, and a good developer managed by a good manager is 100x base developer value. Can you define "good" in a way that would allow us to empirically test this statement that doesn't make this statement a literal tautology?
        6y ago

TameAntelope 6y ago

One question for you: how does an engineer know what to automate?

throwawaygh 6y ago
How does a people manager know what to automate? And, more importantly, what might be the massive unintended consequences of routing your org's permission structure in the same way you route its head-count? Strategic decision-making is not the same as people managing, and people managing can even create blind-spots that get in the way of good decision-making.
- TameAntelope 6y ago
  A manager (who said anything about people manager? I even explicitly said otherwise) knows what to build because she's coordinated with the groups who spend their entire day focusing on the customer, as well as the other developer groups to make sure everyone isn't working on the same thing, as well as leadership so they know what's being built and how it aligns with their goals and vision into the market (because that's what their expertise is in), as well as within her own team to make sure people are building things that can work together smoothly/cleanly, as well as covering for people who have illness/unavailability, and so on... Would you rather have developers wearing every hat in an org? When would they find time to code? I know many developers who don't want to have to do all of that work simultaneously, and are extremely grateful for a manager who can take many of these tasks off their plate. When was the last time you tried to add a recurring meeting to a development team's calendar? They... don't like it, in my experience, and I respect the hell out of that.
  - throwawaygh 6y ago
    Where did I say we don't need managers? I explicitly said otherwise: Invest in your automators and treat your people managers as a low-value-add cost center. Just because you treat something as low-value-add or admit it's a cost center doesn't mean you don't need that work to happen! > When was the last time you tried to add a recurring meeting to a development team's calendar? Yes, I agree, managers' secretarial and political work is incredibly important to the success of an org. But the 1960s idea that it's a "force multiplier" worth high comp is a mistake. I'm not contesting whether managerial work is important or necessary or to be respected. I'm contesting the idea that managerial work is a force multiplier for individual contributors, in any sense other than in the same sense that janitorial work or admin assistant work is a force multiplier. The only orgs where I've seen managers to be actual force multipliers were body shops.
    - TameAntelope 6y ago
      A good manager is worth 10x a good developer, and a good developer managed by a good manager is 100x base developer value. This is why managers often get paid more, and it applies even more the further up the chain you go. A good CEO is worth every single penny of the millions they get paid, because of the multiplicative force they have on everyone below them on the org chart.
      - throwawaygh 6y ago
        > This is why managers often get paid more I think we are disagreeing because we're talking past one another. Hopefully the following observation helps make my point: The average base pay for a manger in the USA is much lower than the average base pay for an engineer in the USA. I'm talking about management as a discipline in general and engineering as a discipline in general. > A good manager is worth 10x a good developer, and a good developer managed by a good manager is 100x base developer value. Can you define "good" in a way that would allow us to empirically test this statement that doesn't make this statement a literal tautology?
        6y ago