Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

People leave managers, not companies

blog.intercom.com

Post reply on HN
Print view View original

161–170 of 186 posts

Re: People leave managers, not companies

#161

Post by glangdale »

ionforce wrote: post #156
glangdale wrote: post #34

Earlier quoted context omitted.

I agree with you that 10X managers exist, but it's utter nonsense to claim that 10X developers don't exist. I've met people who are possibly at the 100X level; it's worth remembering that some tasks cannot be done by an average developer with any amount of time. I imagine that there are probably tasks that don't allow the 10X or 100X to show through; anything basic and repetitive enough (or so heavily specified that…

What's an example of 100X behavior?

100X developers can straightforwardly - sometimes - find solutions to problems that are simple, elegant and fast. This often means that the things that they do are not only cleaner than the solutions produced by weak(er) engineers, but also require less complexity to implement them whether at the organizational level (e.g. because the task is being done by 1 person in 2 hours, rather than 4 people in 200+ person hours, there isn't any time spent on meetings) or at the technical level (e.g. because the task now runs better on a single computer, rather than a cluster or threaded machine, it no longer requires parallelization and a distributed system).

Sometimes this also comes about by a deep knowledge of available algorithms and systems, so while a naive programmer might start flailing around like a dog with their head stuck in a bucket, an experienced person might recognize the problem as an Integer Linear Programming case, etc.

There are plenty of cases where no-one is 100x, of course.

Re: People leave managers, not companies

#162

Post by bcbrown »

maxxxxx wrote: post #93
ZenoArrow wrote: post #79

I think this article misses an important point, which is that managers, or at least line managers, are often the messenger for decisions made in upper management. If I think back on the jobs I've had in the past, it's very rare for me to have issues with line managers. However, I had serious doubts about the competency of upper management in multiple companies that I've worked for. In this case, unless upper manageme…

As manager in the middle you often feel helpless. You want to help your people but you have nothing to offer. Can't give raises, no place to promote people, top management doesn't support your initiatives. It's a difficult place to be in.

You might find this book interesting: https://www.amazon.com/Seeing-Systems-Unlocking-Mysteries-Or.... It's a great view into the differing perspectives of executives, middle managers, and ICs.

Re: People leave managers, not companies

#163

Post by stevarino »

Balgair wrote: post #141
ZenoArrow wrote: post #79

I think this article misses an important point, which is that managers, or at least line managers, are often the messenger for decisions made in upper management. If I think back on the jobs I've had in the past, it's very rare for me to have issues with line managers. However, I had serious doubts about the competency of upper management in multiple companies that I've worked for. In this case, unless upper manageme…

Middle management people are the graphite control rods of a nuclear reactor, they slow down the process enough to be useful and act as sacrificial elements in the power plant. The Gervais Principle is a great lens with which to look at company hierarchies and all people in any large organization should know about it: https://www.ribbonfarm.com/the-gervais-principle/ "Sociopaths, in their own best interests, knowingly…

While I agree with your analogy, graphite is actually a moderator which (in a nuclear context) accelerates reactions. This was a factor for Chernobyl [1] where the rods had graphite tips, so a scram had a momentary rapid increase in reactivity. This burst of reactivity caused temperature to momentarily spike. The heat produced steam which increased pressure and the vessel exploded.

[1] http://www-pub.iaea.org/MTCD/publications/PDF/Pub913e_web.pd...

Re: People leave managers, not companies

#164

Post by Balgair »

stevarino wrote: post #163
Balgair wrote: post #141

Earlier quoted context omitted.

Middle management people are the graphite control rods of a nuclear reactor, they slow down the process enough to be useful and act as sacrificial elements in the power plant. The Gervais Principle is a great lens with which to look at company hierarchies and all people in any large organization should know about it: https://www.ribbonfarm.com/the-gervais-principle/ "Sociopaths, in their own best interests, knowingly…

While I agree with your analogy, graphite is actually a moderator which (in a nuclear context) accelerates reactions. This was a factor for Chernobyl [1] where the rods had graphite tips, so a scram had a momentary rapid increase in reactivity. This burst of reactivity caused temperature to momentarily spike. The heat produced steam which increased pressure and the vessel exploded. [1] http://www-pub.iaea.org/MTCD/pu…

Cool!

Re: People leave managers, not companies

#165

Post by subsubsub »

subsubsub wrote: post #47
crdoconnor wrote: post #43

Earlier quoted context omitted.

IME this kind of experience comes when you have a manager who has no fundamental understanding of what it is that they are managing and has no particular reason to trust you. A manager who can monitor your output by reading your pull requests simply won't engage in this type of behavior whereas a manager who can't will usually instinctively gravitate to terrible metrics like "does he show dedication by being in at 9a…

Is turning up on time for your job a terrible metric? To me being on time is just a very basic low level requirement of being a professional. As the initial commenter said: trust goes both ways. Turning up on time is a good way to show your manager that you can be trusted. Edit: Actually, turning up on time may not make your manager trust you more, but turning up late will definitely make them trust you less.

I wrote the parent post. Here are my reflections on the responses (which were insightful as always).

Regarding turning up on time and professionalism - For me the main component of professionalism isn't trust but respect.

Demanding an employee arrive by a certain time may or may not be a good way to run a business (up for discussion), but, once you as the employee have agreed to do so then consistently being late is a sign of disrespect. Part of how I would define professionalism is doing what you have agreed to do (or making a good faith attempt, even if not possible). If you still think it's unreasonable, consider if your employer paid you a couple of days late - is this still ok?

A lot of the comments mentioned what they consider to be attributes of a good manager, but only from their perspective as someone being managed. To me a bad manager would be someone who allows a centralization of knowledge in one person, so much so that that person can start to behave in a disrespectful manner (YMMV).

Some programmers have the attitude that they are indispensable and can behave anyway they like. There is a certain irony in the fact that we optimize/destroy other peoples jobs for a living but don't consider the possibility that it will end up happening to us - other professions are not as forgiving of some of the behaviours we might consider normal or fair.

Please let me know what you think (lessons on grammar also welcome).

Re: People leave managers, not companies

#167

Post by sdiupIGPWEfh »

cpfohl wrote: post #53
michaelt wrote: post #51

Earlier quoted context omitted.

In software engineering, many companies prioritise "getting your work done" (which is admittedly a subjective, hard-to-quantify metric) and consider any metric that contradicts it to be suspect. A person who believes this would say that if according to policy my best employee should work 9am-5pm, but he prefers to work from 7am-3pm and is my best employee, it is the policy not the employee that needs changing. Person…

This. I've literally never held a job that wasn't okay flex time, and I've worked at 4 radically different types of companies as a software engineer over the last 7 years. I'll be honest, I was surprised to read this sorry if environment still existed for programmers

At every company, my own managers have exorcised far more blatant use of flex time than I've ever needed. If I'd ever received any guff (which I haven't) for my own needs, they would have gotten a deserved ear-full.

Re: People leave managers, not companies

#168

Post by scarface74 »

maxxxxx wrote: post #93
ZenoArrow wrote: post #79

I think this article misses an important point, which is that managers, or at least line managers, are often the messenger for decisions made in upper management. If I think back on the jobs I've had in the past, it's very rare for me to have issues with line managers. However, I had serious doubts about the competency of upper management in multiple companies that I've worked for. In this case, unless upper manageme…

As manager in the middle you often feel helpless. You want to help your people but you have nothing to offer. Can't give raises, no place to promote people, top management doesn't support your initiatives. It's a difficult place to be in.

I bet not all of your peers feel the same way. Some managers are better at playing politics than others and get more resources for their directs.

But I've given up on ever getting a decent raise from a company. I've been prepared to switch companies every two years.

Re: People leave managers, not companies

#169

Post by scarface74 »

Spooky23 wrote: post #133
watwut wrote: post #129

Earlier quoted context omitted.

Showing up sometime between 8 and 11 is unreasonable, because it prevents people to come sooner then at 8. It should have been "sometime between 6 and 11" to be proper. Also since people know the guy in your story is not working enough, it should be easy to check on him couple of times and then fire him.

Our practice is that you can pick your own consistent start time between 7-9. You can also do things like compress your week to 4 10 hour days or add an hour to get a day every two weeks. We don't want you in the office after 6 on anything approaching a regular basis. You can vary the time by day, which is common for childcare or other outside demands on your time. If your habit is a 5 hour start window for an 8 hour…

Is there a reason they need to be in the officd at all?

Re: People leave managers, not companies

#170

Post by yarg »

thrwwyquest123 wrote: post #121
yarg wrote: post #44

Earlier quoted context omitted.

I really don't like the concept of a 10X developer - it implies linear scaling. If you keep writing very similar implementations of very similar things (without actively damaging the code-base) you may very well be an O(n) developer. You keep working at the same pace regardless. If you are a developer of some calibar, the work that you've already done will feed back into the work that you're doing - making you an O(l…

How do I know if I am an O(k^n) developer? My first "software engineer" job I did absolutely nothing but write update scripts, incredibly minute fixes to 10k+ line stored procedures and then helped watch deployments. I left that one for a new one, and I feel very slow compared to my coworker and I know that most of the time my designs for new features are extremely lacking. Should I leave my position since I fear I a…

In my experience the worst developers never ask if if they are doing things wrong and what that might be.

Look at the chages you're making to the code - are they of a quality comparable to what was already there, do they integrate cleanly, are you coding with a style that fits in with the surrounding code?

If not, what could you be doing better and why aren't you already doing it? It may take longer in the short term, but give consideration to your development practices. Eventually you'll find yourself more quickly writing a higher quality of code that's more maintainable with fewer defects.

Post reply on HN