Hacker News Reader

Comment by basisword | Hacker News Reader

I don't understand the shit 'managers' get on here. I've been in this industry for 15+ years and with one or two rare exceptions every manager has been great.

They respect my time, when I need something they're incredibly helpful, and they care about my career development.

IMO the culling over managers over the past few years is really a way to make sure you don't have someone you can discuss career development, promotion, and pay increases with. I have very honest conversations with my managers about these things regularly. If I had to deal with someone a few layers above I doubt I'd have the same success.

Another 'benefit' for the company in culling managers is that the manager track generally has higher pay at each level. Understandable given it seems to involve more time commitment and dealing with people can be much more tricky than dealing with code. Less options for IC's to transition == lower salary burden. Reduce the number of people on the manager track and you reduce the amount of salary an employee can hope to attain. I've definitely been put off switching from IC to manager because I feel the jobs are less secure over the last few years.

View on HN

My experience has generally been that a group of intelligent adults are capable of both planning and steering the course of their development efforts as well as carrying out those development efforts. It's not unprecedented, or a particularly radical thesis: in university research labs, as a PhD student, post-doc, or professor, you manage yourself (PhD students meeting with their advisor once every few weeks).

Sure, there are meta-conversations about process and compensation, and there are younger employees who may need more guidance, and there are intersections with product managers etc. But the ratio of managers to ICs is often higher than needed.

imagine comparing management at a small research lab to a multi-national corporation. Such unfounded hubris.

Imagine actually wanting to work at Amazon

Ahead of the game, soon we'll only be able to spend Bezos Bucks. Company Towns walked so the Company Country could innovate

edit: I've enjoyed two managers since I started working ~2008. The rest were either harmful or generally ineffective, unfortunately. We can get along well... but I know the job's not for me. It's like HR, for the business.

Contrary to GP, career development has yet to meaningfully happen within the same business. I have to move to advance, so I do.

Imagine refuting the idea that people are incapable of self-management.

If you view the university as one large company pushing research forward on many fronts then it is about the same

Universities are not a company model so the comparison is extremely disingenuous.

Have you ever worked at a university?

> in university research labs, as a PhD student, post-doc, or professor, you manage yourself

That is not the real world.

Turns out working for your brother-in-law they let you manage yourself too.

Yes, it is the "real world" for research, of which industry does nearly zero. Research pushes humanity forward. The sort of anti-intellectualism in your comment is part of what is causing the decline we are seeing in society today.

> research, of which industry does nearly zero

wrong and painfully ignorant

I worked in a University lab during grad school, then worked in the private sector for 16 years, and have been back working on research software for the last four years. All I'll say is that the software world should not look at the research world for best practices on delivering software products, except maybe to do the exact opposite of what they do.

I've been in academia for 10 years, now 'out in the real world' for 3 years - I agree with your assessment, the only project management strategy academia knows is 'just work longer hours'.

I have done both and agree

I worked in academia for years before moving to the commercial sector, and in academia management seemed to run on the "in the real world" mantra. Yet if the managers in academia did half of what I saw them do in the commercial sector, they'd get walked out in minutes.

A lot of PI’s pull rank and crack the whip on post docs it’s super toxic and the hours are atrocious with weekends and expectation to work at night. I’ll take a tech middle manager over an arrogant PI every time.

I think it all depends on finding a group of people who share the same goal of making something great together. One person who isn’t interested in that goal can be insidious to a self-managed team. And getting everyone involved means having some reward for doing well, like a validating mission or direct interactions with customers, which can be hard in some roles.

I manage a team of software engineers. While they are all quite good at what they do and care about doing the Right Thing, collectively they're not always great at working towards a common goal.

One of the many challenges I have is that some of them will literally tinker their way to nowhere i.e. they have strong cases of Shiny New Toy Syndrome. If it weren't for me, there would be piles of unfun/unsexy work that never gets done and we'd suffer for it, and it would impact the rest of our engineering org.

It's a thankless job though, I often feel like no one likes me when I'm actually doing my job well. It's OK, I actually agreed to go back into management becauseI was terrified about the prospect of reporting to some new manager my company pulled in off the street (my old manager left).

I'll say this too, while I'm not very hands on these days, I understand what my team is doing and why and can speak with them about the details. I feel like that goes a long way helping me do my job well and understanding what they need to do, to do their jobs well. Non-technical software managers don't really make sense in my worldview.

These don’t even remotely compare. In academia, timelines are long, failure is extremely high, total team involvement on a project is small, motivation is different, as is team selection criteria.

Just look at a large project for academia that requires lots of people and is a deliverable. It reverts to standard practice

I've had mostly bad managers. Most of them maybe wrote code for a year or two and think they understand team dynamics and how to build software. They then burden a good running team with whatever cult processes of the day is without taking any time to understand those team dynamics and which processes fit in those dynamics. It's like a coach that calls nothing but pass plays for a run centric team and makes the 180lb guy play lineman and the 300lb guy play defensive back while thinking 20% turnover is good. No higher understanding of software development what so ever. For me and my teams, they've mostly been a burden.

A good manager protects the team from political shit rolling down hill. They understand who is good at what and allows people to thrive in what they are good at. They keep the team focused, and reward and acknowledge teams for their milestones. They explain to the team why they are doing something and what they hope to achieve while asking the team for their thoughts and adjustments. They also go to bat for the team when it's time for praise, raises and recognition. They privately criticize and publicly praise. They know when a team member is a liability and act accordingly. Most today are just ladder climbers or people who have been Peter principled or nepotism-ed into their role.

I've been in the field for nearly 30 years now. Managers in the late 90s, early 00s were way better than the lot I've experienced since.

Here's a decent summary of how we got here:

https://www.youtube.com/watch?v=P6gMf5zR2c4

> I've had mostly bad managers. Most of them maybe wrote code for a year or two and think they understand team dynamics and how to build software. They then burden a good running team with whatever cult processes of the day

One of the most incompetent women I've ever worked with, a sociopath and pathological liar who to my knowledge never wrote a single line of code, is now a senior manager at Google inflicting pain on some unknown team.

Don't hate the player, hate the game.

>Don't hate the player, hate the game.

Oh, I can hate both. :) The (possibly) good news is now the free money train is gone, companies will actually have to pay attention to how their teams are working. The bad news is they might just chop off heads regardless of ability.

In my experience, most contemporary managers also think they know everything now because they can write genAI prompts without realizing the AI will tend to agree with whatever they put into it.

Micromanagement has gotten really horrible in the past few years. They hire SMEs then discard everything they suggest.

Genuinely unique experience for you probably.

I think career wise in 20 years I'd break down my experience as - 25% benign, 25% malign, 50% good.

This is across 6+ companies, 15-20 managers.

The way people on HN sometimes talk about "management," you'd think the universal experience is 1% good, 9% benign, 90% actively seeking the downfall of civilization

Managers seem like a good example of the "toupee problem" -- the ones you notice, and really remember, are the bad ones; the best, you might never see at all.

I like the old wisdom about apples personally: "One bad apple can spoil the whole bunch".

The economic cards seemed more stacked against workers, so resentment for managers builds more quickly than the opposite direction. Management who resents their workers can also fire them. Workers who resent their managers must go find another job. And offshoring/nearshoring for workers happens more than managers.

Would be interested in comparing the interview processes for ICs vs. managers at Amazon. Probably no leetcode-equivalent for managers?

A lot of women complained about the "boys only club" in the 70s/80s/90s. While things have gotten better for women, there is still a lot of "club" going on.

> "boys only club" in the 70s/80s/90s

> things have gotten better

If you read up on humanity, you'll likely find that it would be a bad idea to put money on any wager that this will end.

> Probably no leetcode-equivalent for managers?

For FAANG, most front-line manager roles have a leetcode style coding round, though it may be easier and/or weighted less heavily than for an IC.

Maybe it's because this isn't about looking at other people's toupees?

At the least a bad manager can make the place you spend a majority of your waking hours a worse place to be, and at their worse they can permanently harm the trajectory of your career or even your mental (and by proxy physical) health.

Some analogies are limited by the weight their original context conveys. I wouldn't let a surgeon get away with "the ethics board will only talk about this heinous thing I did, when most of the time they don't see me at all".

I think you also need to account for the amount of bad.

Like a bad programmer can push terrible code, get caught at review and performance managed.

But a bad manager can cause much more harm. An organisation with bad management can punch itself in the face very hard and cause significant issues.

Like I have only seen a terrible management culture in 2 of my employees, but for 1 of them, it lead to:

30 or 40 careers damaged, internal stalinist purges. Months wasted on drama. 21 million yearly recurring in wasted IT expense. Probably close on 500 million in non recurring waste over 4 years. 4 million yearly recurring in executive waste. Significant brand damage, significant resume damage for people who worked through it. Actual end user harm.

> Like a bad programmer can push terrible code, get caught at review and performance managed. But a bad manager can cause much more harm. An organisation with bad management can punch itself in the face very hard and cause significant issues.

Ironically every place I've worked, a lot of these bad programmers got placed into the management pipeline, because they had not the skills to hack it as an IC, so it was a worst case scenario of "fake it till you make it".

You could be the most incompetent programmer in the world, but a suave bullshitter is a shoe-in for management, where they now get to tell the competent programmers how to do their jobs.

Yes managerial blast radius is an order of magnitude larger, and ability to measure performance lower / more lagged.

In all fairness, a bad programmer can do a lot of damage. They can create attack vectors, destroy data and cause production outages. None of that is good. There are things in place to prevent that, but if those also fail, it'll be a busy week. Ask Crowdstrike.

https://ezo.io/blog/crowdstrike-outage-and-the-blue-screen-o...

Bad managers can also lead to a lot of bad programmers because all the good ones left. I reckon a bad manager is a multiplier of bad programmers. Bad is just bad no matter what level you are in.

I think the main difference is that programmers have (or should have) more transparent processes that they go through (code reviews, design doc reviews / signoffs, CI/CD to catch mistakes) compared to managers. Granted, I'm an IC so I know more about those - maybe managerial roles have similar ways to ensure you catch the "bad" early enough.

I think its genuinely harder to measure the squishy productivity / collaboration / delivery output stuff that managers are supposed to own, but also .. managers aren't as incentivized to create processes to measure managers ..

Fair, however if a bad programmer isnt subject to code review, thats sort of a management issue.

There's a reason why the saying goes "people don't quit companies, they quit bosses".

I don't think the good ones go unnoticed.

You remember what you learned from good managers, and you remember how bad managers made you feel. Benign ones could be replaced by an LLM.

>>you'd think the universal experience is 1% good, 9% benign, 90% actively seeking the downfall of civilization

It might not be a grand conspiracy or might not come from meticulous planning, but what happens is they just work for self preservation. Its no secret that people who work on a thing are bound to know things about them better than some body who just approves leaves, or makes abstract decisions. You will get replaced if you don't assert authority often and proactively kill the biggest threat to your position. This also means maintaining pets, and rewarding them more than people who are performing better.

All of this resembles a pattern of behaviour over the years with managers sabotaging everything good around to save themselves.

Over years I have seen managers are the biggest reason why companies go down. There are few other reasons.

Right and there's asymmetry between what a good & bad manager can do.

A good manager can motivate people, but in absence of interesting projects or budget to pay people, they can still lose talent.

A bad manager can make even interesting work and good compensation intolerable and lead to a revolving door of B players as no one with better options wants to stick around.

The through line for me of genuinely bad managers is guys (always guys) who are emotionally unstable and/or have anger management issues. Whether it was being put on as an effect to get a result, or they genuinely had no control, it doesn't matter.

Shouting at people, berating people in public, praising in private while criticizing in team meetings, going on tirades over petty stuff (formatting of internal non-user facing documentation), etc.

It's just negativity bias. We're hardwired to remember our wounds and avoid getting them again. And apparently we're hardwired to engage with those more than praising the great ones.

For some harder numbers though: it seems to also follow the pareto principle : https://www.pewresearch.org/short-reads/2023/11/13/most-amer...

20% of "fair/poor" management ruining the 80% of good.

(though this is from 2023. A lot has changed in sentiments since then).

For a lot of people, the difference between a good manager and a benign manager is smaller than the difference from benign to bad.

If you are working for an emotionally unstable shouter, the negativity seeps into your work and quality of life. I've seen bad managers make adult men cry (rarely for any good reason) and still last 2-3 years.

If you work for a good manager, you learn a little more stuff, maybe.

Do the break down for developers.

A manager can put you on PIP, have you fired, and make your life miserable for months to years. What can an engineer do? Write some bad code that is a little annoying to refactor?

In finance a bad engineer can bankrupt the firm.

If a bad engineer bankrupts the firm, it's much more the fault of many others.

A finance institution with no code review processes is one deserving to go bankrupt.

I'd love to hear how a single IC can do this.

Developers get marked to market quite quickly in annual reviews, if not sooner. Junk PRs, bad code, tons of bugs, acting like a jerk - it catches up fast. I've seen devs walked out the door in first 90 days of probationary period, or cut in their first annual review.

For a manager there is a longer leash as the things they can impact are harder to measure with long and variable lags. So it can take 3 years easily for an obviously bad manager to be dealt with.

My career hasn't been as long but it's been 50% good 50% bad roughly - of the bad ones, one was a sycophant to a narcissistic product owner but didn't directly cause me any trouble, while the other two were promoted developers who tried to force their will on developers while also playing political games to try and preserve themselves.

Unique experience. I've had mostly useless managers in my 15y career, downright toxic ones in my 3y AWS stint.

What makes your experience representative and theirs unique? How are you measuring this?

You chose to work in a toxic company, of course you had toxic managers.

There was that comic about org charts a decade ago:

https://bonkersworld.net/organizational-charts

The reality is that the Microsoft style of organization is very prominent in the industry.

You don't quit companies, you quit managers. I've fortunately had great ones who balanced being into my development with making sure the job was being done. But that's not a universal experience, sadly. I work in games so you can find plenty of horror stories on what happens under bad management without me giving second hand stories of other teams I worked next to being raked in the coals.

I simply want to focus on working "in the ground" so management never really rung for me. My endgame goals focus on the opposite of managing a large beauracracy of tech workers on a massive project.

Look to give you the benefit of the doubt there are some experiences you probably just havent had.

I've also had all good managers who helped me move up levels, etc. across 3 companies for about 5 years. I think you're right about the side effects, and it's sad to see. It reflects, in my mind, an acknowledgment from these companies that they won't grow as much in the future as people right now think they will

One or two rare exceptions over 15 years must account for several bad years, no?

As for giving and getting shit, if you evenly distribute matches over a quadrant of good and bad dev-manager pairings, then 3 out of 4 matches are gonna have a bad time. And even in the top 25% where both devs and managers are good, you can still have a personality mismatch, or other troublesome contextual factors, like your boss's boss, the company's success, the head winds, etc. Work relationships can strengthen or crumble under strain. Of course, the best way to maximize your odds is to do your best to be a good person to work with, but the odds are simply not in your favor in the first place for either party.

I’ve experienced so many kinds in 12 years including those who were incredible to work with and those who contributed nothing, and a bunch in between.

A lot of managers can be automated. Few are making creative decisions that can be replaced by a computer, and smaller teams of managers.

The primary output of managers is not "making creative decisions" in fact that's probably not anywhere close to the top line.

And because they are only doing administration and reviews, they are even lesser required.

Yes, you can always push that admin and coordination and communication work down to the ICs. I'm sure you'll be thrilled to get that added to your plate.

Are you talking about the same admin and coordination required to keep 10 layers of management updated?

Easy enough to solve, just remove a few layers.

Sure, and while if you want it to stick, you'll have to get rid of the CEO and the shareholders, too.

Those 10 layers of management work for them, not you. That's why they exist.

Yes, and that's why they are removing the 10 layers. It's not working for them any longer either.

They are trimming one of them. But the work will likely get pushed down.

And a large amount of that work will disappear because a significant amount of that work is merely maintaining reporting and answering to bosses ad-hoc.

Removal of bosses is literally more efficient because it reduces the amount of work per unit person.

They aren't removing your boss, you still have one, except your new boss (who was your former skip level) will expect all the outputs that your previous boss produced.

From you.

> They aren't removing your boss, you still have one, except your new boss (who was your former skip level) will expect all the outputs that your previous boss produced.

But you do realize that previously, you had to produce outputs for your boss and skip - but now, you only need to produce for skip?

For 2 bosses, the hypothetical low/mid manager has to consider 2 bosses perspectives, 2 bosses politics, 2 bosses promos, 2 bosses peer teams, 2 bosses meetings, 2 bosses technical skill, 2 bosses status reports, 2 bosses committees, 2 bosses ad-hoc demands.

With 1 boss, that work reduces to 1 bosses perspectives, 1 bosses politics, 1 bosses promos, 1 bosses peer teams, 1 bosses meetings, 1 bosses technical skill, 1 bosses status reports, 1 bosses ad-hoc demands.

Removal of 1 boss led to 50% decline in work for 1 low/mid manager. Scale that to 10 low/mid managers, you now have a large amount of time freed up - SIMPLY because 1 single org chart bottleneck has been removed.

If your boss was any good at his job, he did most of the reporting and communication to keep the skip level happy.

Either he wasn't, or you are just not aware of all the work he was doing.

A good manager is a shit umbrella. Most of my managers have been good.

> If your boss was any good at his job, he did most of the reporting and communication to keep the skip level happy.

A big IF - and now the company has to hire more HR and more bosses to check on this boss. Create more bureaucracy around checking "if they are any good".

> did most of the reporting and communication to keep the skip level happy

And someone had to do reporting and communication to this boss in the first place. Thus, a boss requires someone/some people below them to spend time reporting to them, and themselves needs to spend time reporting above them to the skip.

From the big bosses perspective, this miniature boss's existence added more work to the company than required - thus, leading to the original post.

Just yanking layers to minimal required bosses is enough to reduce total work.

Yeah, I’ve been working in the industry about 20 years, and I’ve had one bad manager experience (and also a couple of not-great experiences where I didn’t really _have_ a manager).

Tbh I think this “we should have fewer managers” thing is just the current management fad. It’ll pass sooner or later.

> IMO the culling over managers over the past few years is really a way to make sure you don't have someone you can discuss career development, promotion, and pay increases with

That’s the point, surely

You should count your lucky stars.

I had two good managers, the rest ranged from innocuous to malevolent. One manager even cursed me for refusing to approve an engineering deviation to allow a passenger plane to fly when the wing composite was delaminating. He said he went through the trouble of preparing the pseudo legal document and how dare I refuse to sign. I told him a) I was not the SME on wings as I was an engine guy b) if this was such a no brainer why didn’t he or the SME sign their approval. This was when I worked at a major airline and wasn’t the only egregious thing I had experienced. This incident was one of the reasons I switched to IT because in software it was unlikely you could be criminally held responsible for such irresponsible behavior.

Anyways. My guess is if I had signed off the pilot would not have accepted the plane during his preflight. Then an investigation would have started on who would have signed off on such a thing.

Reply on news.ycombinator.com