First, Let's Fire All The Managers | Hacker News Reader

First, Let's Fire All The Managers

archive.harvardbusiness.org

118 pointsvellum79 comments

From what I recall, there's a big counter-argument to this, which is that management is necessary because organization is fundamentally necessary, and that when there are no formal "managers", certain people wind up being informal managers, but because this is "hidden", it winds up being less accountable, more open to cliquishness and/or discrimination, and having a host of problems of its own.

I've been trying to find an article I remember which describes how this is particularly bad for women, who can easily be "informally" excluded, but I can't find it... All I can find is this, on Valve:

http://www.pcgamer.com/2013/07/08/valves-flat-structure-cont...

I think the article you're referring to is called The Tyranny of Structurelessness.

Article: http://www.jofreeman.com/joreen/tyranny.htm

Wikipedia page: http://en.wikipedia.org/wiki/The_Tyranny_of_Structurelessnes...

This is a wonderful article, and details quite nicely why you can't simply throw managerial theory out the window and pretend everything will be okay. It goes into depth about the problems one encounters if an organization attempts to eschew all management, and then goes into depth about how to fix these issues while maintaining a flat organization (in short, just because you acknowledge that a managerial position exists does not mean that that position needs to hold asymmetric power over the people who are managed, as occurs in the traditional hierarchical model).

Indeed. Doing away with the 'formal' structure doesn't mean the function of that structure also goes away (if that makes sense). This is true for many things e.g even if you don't have a sales team in your company you still have a sales function. Superficially getting rid of something (say, 'management') instead of dealing with the specific problems (e.g. miscommunication, poor incentives, opaque decision-making etc) doesn't seem like a long term solution, though it may be very satisfying for a couple of years.

Something I found ironic when I read that article is that it (very nicely) articulated things I've known since school, just by observing social interactions. I suspect most people also know these things but they get forgotten as we grow up.

When Ryan Carson's post made the rounds here not too long ago, I was really surprised by what got them thinking about removing their managers. http://ryancarson.com/post/61562761297/no-managers-why-we-re...

  By 2013 we had grown to 60 people with seven managers and four
  executives. As we added more people to the team, we noticed
  something disconcerting: rumors, politics and complaints started
  appearing.

If you have more than two people in an organization of any kind (company, sports team, high school, PTA), you're going to have all of those things, and it has nothing to do with good or bad management. They seem pretty happy with their decision to have changed their structure, but if they think that the rumors, back channeling, and quiet lobbying are gone, I think they're going to be disappointed.

I don't believe having managers is bad, I believe in having too many is what really causes issues. I have seem some organization charts where managers managed a single person and they themselves might be the only report to another. Its like the title is giving out in lieu of compensation or to not hurt someone's feelings. Worse is assigning people to someone just because their a manager regardless of the assigned person's work.

So what is the ideal proportion of managers to workers - and do you count a manager who gets his/her hands dirty to be a worker or manager?

So I spent some time in anarchist circles, and what you've just described comes up regularly. It can actually be even worse than what you just described, specifically because the informal managers are also not given the social status a formal manager is. That means that in worst case scenarios and poorly working teams, they are expected to both keep up the workload of the position they officially hold, while also keeping up the managerial position (burnout), or alternatively they are regularly challenged as to the validity of their managerial knowledge.

That said, in rapidly changing situations, that second component is actually a great strength, as it ensures a free market of ideas within the potential "manager" pool, and allows an organization extremely rapid ability to respond to change. And so long as there is some intra-team awareness and team focus on measurable goals, the former can be placated as well.

There is another weakness, in situations where the managerial overhead is extremely intensive, unchanging and time consuming, ie, when the situation results in people who informally spend all of their time doing management, simply because it is more efficient for one person to do it, than that burden to be shared on everyone. The positive to this is because the position is informal, the number of people working on it will be exactly proportional to the amount of work required to be done, while in a normal organization there is a drive to increase management as much as is sustainable by the organization, both as employees attempt to gain higher salaries, and individuals already in management attempt to retain theirs. The negative is that as the informal position moves between individuals willing to put in the time to do the work, if there is no process to track who has done what, accountability can be an issue.

To attack this accountability issue, the traditional anarchist response has always been to formally recognize these positions, and work hard to ensure that these people are replaced regularly by democratic process. Someone below linked the tyranny of structure-less management article, which goes into far more depth, and arrives at the previous sentence as its main conclusion.

In our modern software startups, these are exactly the situations we attempt to automate, so it is less of an issue.

So now we are left with an issue of the size of the organization which wants to attempt this managerial structure.

In small groups, such stringent processes as just described need not be defined. If you have a decently working team, this is how people coordinate by default nature, and this informal position will move between people as different individuals show initiative. A simple system of ledger books or signing line on various forms will likely be sufficient to keep organized most tasks undertaken informally.

If these positions are not acknowledged, it begins to break down once you reach teams where members are counted in 3 digits, hundreds or several hundreds, simply because now you have to recognize that individuals do not work in groups of such size. So really, you are working with a large group of groups, and the dynamics change quickly. Unfortunately, I cannot comment on how this works in reality. In modern America, by the time a group grows to this size, it is considered a political threat, and ripe for extrajudicial dismemberment by the state. I do not know how it would act in the comparatively safe environment of corporate organization, though I imagine most of the problems could be alleviated by automation and programming, such as a good CMS system, but I have no experience there.

> "... the number of people working on it will be exactly proportional to the amount of work required to be done … "

I strongly disagree. This is a recipe for burnout because it ends up being done by the people who care most and are willing to sacrifice more than others (in terms of time/health/whatever). Taking the view that 'the work got done, therefore this was the correct number of people' does nothing to account for the human cost. Example: One person can organise and run a conference with hundreds of attendees and tens of speakers but I'd argue it's not a good idea. FWIW I've seen team sizes between 1 and 75 for exactly the above scenario.

I've been in this kind of position (great learning experience) and seen others go through it but it's a completely unsustainable way to run any kind of organisation.

It seems like this is becoming more and more of a trend, and is just starting to push into the realm of legitimacy, rather than "kooky company does something crazy".

There's plenty of examples in tech of companies without traditional management structures (and mostly no management at all) - GitHub, Valve, Treehouse, etc. Now a food processor of all places pulls this off.

I do think this is an approach that doesn't necessarily work well for large companies - building the sense of camaraderie necessary for this is impossible at 10,000 or even 1,000 employees IMO, but on the other hand - who says we need to have giant companies? Wouldn't things work a lot better with a whole lot of 150 person companies?

This article is from the December 2011 issue, it's 2 years old now. However, the self-management principle at Morning Star still seems to be in effect: http://morningstarco.com/index.cgi?Page=Self-Management

WL Gore & Associates has over 10 000 employees and operates on a similar structure (and has done for decades):

http://en.wikipedia.org/wiki/W._L._Gore_and_Associates#Cultu...

I recall an interview with John Chambers of Cisco many years ago where he mentioned having 20 or 30 direct reports, to avoid too much meddling in people's affairs. This can work if you don't need to make drastic changes (we have had a massive disruption in our industry and need to let half the workforce go - which half?) and if there doesn't need to be any cross-functional tiebreakers.

What's strange is in a more recent (2009) article [1] he claims to be more command and control, but shifting towards a more bottom-up organic method.

[1] http://www.mckinsey.com/insights/high_tech_telecoms_internet...

Today JC's direct report more or less == Cisco Board.

He really only has three: COO and Co-President Gary Moore, Co-President Sales & Development Rob Lloyd, and CFO Frank Calderoni. Pretty much all other C-level and EVPs report to Gary and Rob.

John talked about cutting, primarily, middle management during the recent 4000 job cuts announcement. You can guess how well that has worked. There is still plenty of scope to broaden the span of control in Cisco.

A similar but more fringe structure that is regaining popularity is the worker collective/cooperative. Mondragon pulls it off to some degree with 10,000+ employees but their structure definitely involves more hierarchy than the smaller coops. 10,000 employees doesn't have to mean 10,000 employees all doing the same work in the same division, you can still have camaraderie within divisions or work groups and organize the entire company democratically. The worker collective gives ownership and influence to every worker and lessens the "us vs. them" dynamic by empowering workers and basically eliminating managers.

I thought Valve didn't pull it off once it got to a certain size? http://www.wired.co.uk/news/archive/2013-07/09/valve-managem...

Yeah, stuff like that is what I was getting at about it not working for larger companies.

It seems Ellsworth's firing was hyped a lot around the tech news, and she definitely leveraged that to get her new AR goggle project off the ground.

I'd like to see this story corroborated by other former Valve employees. Her interview there feels a lot like self-victimization.

The Brazilian Semco company has a similar structure, IIRC:

http://www.schneede.se/assets/files/Ricardo_Semler.pdf

I'm somewhat surprised that everyone is so anti-management here.

First of all, Morning Star seems like a pretty well-run place, but the whole system works because they have the luxury of hiring the right people. This is, of course, the single most important factor in any effective organization. If you don't hire the right people or play to their strengths you won't be able to pull this off.

Anyway, a lot of the reasons listed in this piece look like they're sourced from companies with crappy management and I don't think they represent the reality of management, at least in a lot of places.

At Facebook (the only company where I have experience as both an IC and manager) there are three important things to know about management:

  * It's a parallel track to IC (individual contributor aka hacker). You aren't "promoted" to manager. It's a lateral move. There are plenty of ICs making more money than managers.
  * All managers go through "bootcamp" (6-week engineering onboarding) and commit code.
  * The role of the manager is to basically do all the shitty work so ICs don't have to.

I want to stress that third point. If you're anything like me, as a hacker you want to be given a well-defined problem (unless the problem is defining the problem) and go off and hack on it for a while with minimal distraction. There are a lot of crappy distractions that come with working in the real world. Here are a few:

  * Dealing with interpersonal issues
  * Onboarding new hires
  * Coordinating engineering efforts and priorities (aka: going to meetings so ICs don't have to)
  * Telling that guy in sales or that gal in product management that we can't work on their tasks right now
  * Dealing with medium-priority bugs during crunch time

If I'm shipping a product I don't want to deal with any of this stuff as long as I'm not being micromanaged or feel like a second-class citizen to management.

So what managers end up doing is being responsible for all of the shit work so their team can go do what they do best. I'm actually concerned that this whole "no management" movement is going to lead to less motivated and less coordinated teams as these companies grow.

> as a hacker you want to be given a well-defined problem

I think identifying and solving open problems is more of a hacker mindset.

The problem is that menial work (arranging meetings, organizing schedules, handling communications with colleagues and clients by phone and mail) used to be handled by secretaries. If Mad Men is to be believed, at a certain employment level you would get a personal secretary and below that you could draw from a common pool. Where did all these secretarial jobs go? Their title was changed to "manager" and they were put in charge. God help us all.

Um, no. Secretaries were replaced with personal computers and websites. Nobody needs to take dictation anymore, after all, and information is now at everyone's fingertips.

And the logistics of arranging meetings etc. are handled by office managers generally, if you're talking about who uses which room and for how long.

The idea that secretaries turned into management is bizarre.

I think this is a generational gap. As far as I am aware their is a gap between what the role of secretaries was seen as (typing, filing) and what the range of roles contained within the post actually did. There was a good reason why even now, some people said that a computer, and great search hasnt caught up with an excellent secretary.

I know of dozens of secretaries and typists in the medical field. Voice dictation software doesn't cut it yet. Software that works sort of ok is dependent on platforms companies don't have, and humans are better at error correction, answering the phone and resolving problems. Many surgeons seem to have 2 or 3 secretaries to keep on top of the work. It might be an industry or country specific thing, but secretaries aren't going anywhere fast in my little world.

So what managers end up doing is being responsible for all of the shit work so their team can go do what they do best.

This is the definition of management if it is being done correctly. In effect you manage inbound tasks from upper management to ensure that important tasks get completed while FUD is diverted or eliminated. This strikes at a larger point though that if you didn't get crap from above there would be little need for management.

The issue is when everyone thinks their project is the most important and should have resources allocated towards it. Then you need someone with a better strategic picture to marry all of the competing requirements such that it fits into the position that your organization takes in the market for that time period.

It also reinforces the fact that there are different layers of "knowledge" from strategic global markets to lines of code. If you want anything better than luck to line those two things up across dozens or hundreds of projects you will need some kind of middle management.

Wow, those three dot points sound like dream-land compared to my current company.

I think when people say "get rid of managers", they are not referring to the setup you mention, but to the old-world where managers are clueless about the work actually being done, take all the credit, and make much more money. Also, being "promoted" to management means leaving tech forever.

I have found the approach like the Facebook example above to be more common in the UK, to South Africa or India. I have been wondering whether it's just my experience, or if developing countries are just simply behind and resemble an old class system. That said, in all these countries a specialist surgeon would be seen as an equal or superior to a hospital manager, almost exactly like the Facebook example. So, the issue may be down to the immaturity of the IT industry.

Management practices, on average (there are always exceptions) tend to be much better in the developed world. Here's some research from Harvard Business School that looks at it more quantitatively: http://www.hbs.edu/faculty/Publication%20Files/12-052.pdf Here's another one, although it's less academic and more consultanty: http://www.stanford.edu/~nbloom/ManagementReport.pdf

Improved management is actually one of the reasons that developing countries are able to achieve such remarkable productivity gains. Here's a study that actually used Randomized Controlled Trials to show the effects of improved management practices in India: http://www.stanford.edu/~nbloom/DMM.pdf

The TL/DR on that last one is that a small amount of management consulting led to pretty big productivity improvements.

I've spent much of my life in a poor, developing country, and in my experience there is a huge difference in how people perceive formal structures in general, but power dynamics in particular. People in charge tend to be more authoritarian, and people in lower positions tend to be more subservient and they do exactly as they're told and no more than that.

Except for the ones who have a really obvious stake in the structure's success (owners usually), the thing people in each of these positions have in common is that they do not perceive themselves as individual, autonomous actors with responsibilities, but rather as roles to inhabit and strictly follow - and never go above and beyond.

This often led to hilarious and infuriatingly inefficient results, and me and my family constantly marvelled at this, even after ten years in such an environment. Some examples:

In construction, the boss would have to be on-site at all times. He'd just be standing there watching his workers work, because if he left, they would immediately slack off. And if he gave instructions, they would follow these instructions to the letter, sometimes with disastrous consequences. When asked why they did something particularly stupid, they'd shrug and explain that they were told to do this and so they just did it. It often had little to do with their abilities.

When an individual worker would be promoted to a managerial position, he or she, no matter their personality, would instantly turn into a rather unpleasant bossy character, and not take any initiatives beyond 'managing'. Working alongside their underlings was unacceptable. Seeing this sudden transformation was a bit scary.

A translator working for my father wasn't very familiar with computers but when tasked to translate some text using word perfect, she went ahead and started working hard. A few hours in, he checked up on her and while she said everything was fine, he noticed that she hadn't translated more than a page. After some prodding, she explained that she was rather frustrated that as she reached the end of the page, her earlier translations vanished. It turned out that she had translated a screen full of text multiple times and simply didn't understand that off-screen text was still there. At no point did she feel an urge to ask about this, despite her frustration.

I could keep going, but what really struck me, again and again, was the sense of lemming-like behavior that people exhibited as soon as they were assigned a role, the huge inefficiencies of having multiple people do very specific jobs that one person could do faster and better in the same amount of time, and the complete lack of a sense of 'professional pride' in their roles.

What was especially interesting about all this is that my 'home' culture (Holland) is pretty much on the extreme other end of the scale (often, not always): bosses dress down and tried to avoid any kind of pulling rank (to a fault sometimes), employees have opinions on everything, even things outside their responsibility, and they express these opinions vocally and often with little regard of rank and appropriateness. Flat power structures (polder model) are favored over big hierarchies.

When children in this developing nation talked back to their parents, they'd receive punishment whether they were correct or not. When Dutch kids talk back we consider them cute and precocious.

These are broad strokes, of course, but these differences were apparent time and time again.

I have some theories as to why this is the case, but none that seem to fully explain these massive differences.

>> "What managers end up doing is being responsible for all of the shit work so their team can go do what they do best"

This is, in essence, a manager's most important job. It gets under valued, because when it's working it doesn't look like a manager is doing anything, it just looks like a bunch of great people working away. You notice managers when they're bad, not when they're good.

Sorry, but IC is what?

Individual Contributor i.e. someone that doesn't manage others, but contributes as an individual.

Disclaimer: I work for a Fortune 500 company. I suspect the inefficiency added by managers is overblown, because most of them do not spend 100% of their time on traditional management activities such as supervision and decision making. A lot of their work is on tasks that would fall in somebody's lap within any organization. Some of those tasks are ones that I find to be dreadful.

Most managers will instinctively delegate dreadful tasks. That's part of the fun of management.

I wonder if there's any managers around who also code? Maybe 50% of their time could be management, 50% coding. Nothing like getting hands dirty with the company code to see where things are at.

Shared management roles could also be an option. Or, manage by committee? There's enough design by committee happening, surely we can all jump in on management decisions?

> I wonder if there's any managers around who also code? Maybe 50% of their time could be management, 50% coding. Nothing like getting hands dirty with the company code to see where things are at.

An important part of the reason why so many technical people have an "overblown" idea about the inefficiency added by management is negative experience.

I have met maybe a dozen people who were in management position and were not completely and utterly incompetent at what they were managing. More often than not, the managers I meet were people who started out as programmers, but quickly realized they don't stand a chance building a career out of it. This is an incredibly common career path.

This. I'm working as a consultant, and I see the same thing over and over again. There aren't many "IT Managers" that are competent at what they're doing. If you are good at what you are doing (which might be programming), go and look for an environment with like-minded individuals with a sense of craftsmanship and pride in their work. Everything else will suck the life out of you.

> Most managers will instinctively delegate dreadful tasks. That's part of the fun of management.

I think this becomes one of the reasons that so many managers end up being viewed in a negative light.

Surely part of a manager's job is to protect the team from external pressures. The decision to delegate should be made on the basis of the impact of delegating / not delegating, not on how dreadful the task is.

As one of my old (properly self-deprecating) managers used to say "If you're a lightweight you'll float to the top eventually."

The best managers I've had are ex-coders who have fully realized they cannot both code and do a good job as a manager.

I had one really good coder cow-orker who became my manager. He tried, and it was pretty poignant at times, but eventually realized the truth. Until he did, he mostly sucked at both jobs.

Yeah, this pretty much sums up what I've seen as well. I know good ex-coders who became good managers, but the stepping stone to doing it was having to stop writing code.

I've tried. It doesn't really work. Organizing is a "heads up" activity and trying to achieve flow while being available to answer questions and stuff is frustrating at best.

This is my life right now. About a month ago I took on ~4 reports, but still need to maintain my previous productivity levels. It's really hard to both be available for them / our client and still get into my groove.

> I wonder if there's any managers around who also code? Maybe 50% of their time could be management, 50% coding. Nothing like getting hands dirty with the company code to see where things are at.

Yes.

Though there is such a mental disparity between coding and higher level management that it is not 100% efficient.

You're either coding or you're managing. You can't be knee deep in an obscure rendering bug whilst also still mentally committing to keeping track of what everyone else is doing. Getting back up to speed (from either one to the other) always takes a small amount of time.

I've worked at places with this, usually called something like rotating team lead, or duty engineer, or ops duty, or pager week, or something like that. To the best of my knowledge everyone involved dreaded when their week came up. I've only experienced this implemented WRT (metaphorical) fire fighting not traditional dev work. If an earthshattering bug appears this week, you'll be fixing it and/or triaging it and/or figuring out who can help you fix it.

Its awesome at getting rid of "pass the buck" and awesome for everyone else in the org to have a single point of contact (even if its a different human each week). Not so awesome around changeover time or if your docs aren't up to spec or when you discover the hard way someones code isn't good. Also not so hot at fixing major architectural issues (like, its going to take more than a week to fix this correctly, and I'm only on duty for a week, so whip out the band aids...)

Rotating through on call, especially having not just ops, but engineering rotate, is a huge benefit.

As much as DevOps is being promoted, there are still two fundamentally different skillsets and job descriptions, and if its just the Ops side of the house getting the 2am wake-up pages, you're going to get a decidedly different attitude toward Things Wot Breaks Prod than if the developers are on that rotation as well.

Plus it means rotation comes up that much less often, which is a good thing.

Parkinson's law and Peter's Principle apply to privately held corporate structures as well. There's whole mountains of busywork and paperwork related to hierarchical structures that give the impression of "work" to both outsiders and insiders where it's really just non-productive paper shovelling.

https://en.wikipedia.org/wiki/Peter_Principle

https://en.wikipedia.org/wiki/Parkinson%27s_law

Mind you, I'm not sure getting rid of the managerial staff is enough to address these problems fully.

This is the best comment.

What sort of dreadful tasks are you talking about?

There are plenty of things that managers do, that could just as well be done by some kind of junior administrator.

Tasks that require someone to be fairly articulate and numerate, and at a level where they can be "privy" to sensitive information. From a technical standpoint, a lot of it could be described as data gathering, analysis, and reporting, in situations where those things can't quite be automated.

Comment above about Parkinson's Law and the Peter Principle are duly noted. ;-) I was a manager for a while, so I don't exempt myself from those observations.

I have seen this too. I worked for a while as a contractor for the U.S. federal government (IT/programming), and saw that the next rung up for me within the fed was becoming a "full fed" and entering management. I saw what many of those people did and ran away screaming, taking a higher-paying alternative coding gig instead.

One of them had worked in the past for Microsoft and said the fed was remarkably similar. I suspect any organization approaches the fed exponentially as its size approaches, say, tens of thousands.

I've seen lots of these "x-role free" companies: no managers, no sales department, etc. Inevitably the function that those roles take on end up on somebody. Companies will sit in denial about it for a very long time, even making up weird titles to pretend like they still don't have managers, but ultimately they'll end up with them in the end because that's how they have to interface with the rest of the world and many of these roles exist because it's how work naturally breaks down and how people naturally specialize.

I'm extremely suspicious of the word "naturally" there. What nature are they conforming to?

For example, there's a natural split between people who need to spend their days quietly writing software and the constant interactions and distractions of sales. It's almost mentally impossible to do both, not because either job is hard, but the context switching means software never gets developed and when writing software, possible customers never get called and deals never get closed.

The point I took from the article was that it's better for managerial tasks to be distributed across the team than reside with specific individuals. Of course there's no free lunch - an self-managed coder can get less coding done than a managed one, because they have to spend some of their time on managerial tasks. But it's possible they'll be more effective at it than an external manager.

I think in some organizations this should happen anyway by giving the individual wide degrees of autonomy. Micromanagers are the bane of everybody's existence.

The file name -- [B@58aa11eb.pdf -- seems like it was created by something written in Java: instead of printing the real file name, they printed the signature of the Byte array holding the file name data.

See here [1] for a better-than-Oracle explanation:

[1] http://stackoverflow.com/questions/1040868/java-syntax-and-m...

I'll try a pitiful attempt at a TLDR, feel free to try an improve it (if you can):

The IRS has extremely strong opinions on what it means legally to be a contractor. However you can easily avoid the legal issues and run a company as if it were full of contractors, if you want, by merely treating them legally as employees but treat them managerially (sorta) as contractors. And at least anecdotally sometimes this works really well.

Or a really short TLDR is contemplate the building trades and the role of a general contractor, and run your (probably) non construction company that way.

Umm, my take is that developers are taking over the world, in term of duties. We don't need BA, since we do agile development and talk directly to customers. We don't need DBA now that we have ORM and NOSQL. We don't need sys-admin/sysops because we are devops now. Now, no managers. Next, no VC or no boss?

HA! This might work for smaller companies. Try doing this we a 25,000+ FTE plus 5,000 or so temps and contractors in a company with a presence in around 100 countries. Although our managers dont supervise us constantly they have their own work to do. They usually give us tasks or projects and then we give them updates.

This really needs an NSFW tag...

I like how that can be interpreted in more than one way.

Why?

It's a joke. Normally NSFW tags are for porn-ish links, but this is unique in that it's not porn but could potentially make your manager equally upset.

Mediation sounds good, and the 6 member panel idea also good.

In online tech, it pays to keep the mood relaxed and the office comfortable and spacious. Regardless of how much "serious business" is happening online, the internet is still a chilled out place.

The traditional hierarchy of managers is inherently unrelaxed. When you have multiple managers surrounding you, their presence might cause some to hold back on decisions, or pause that initiative. The managers will take care of all those nasty little details such as making decisions, you just keep doing the task you were delegated.

Remember what it felt like when the teacher left the room? If the manager leaves the office for the day and you feel that same rush of freedom feeling, you know things might be better without managers (or just a new manager).

Somehow, I doubt this will be an easy sell to the management hierarchy.

Horray! Stick it to the man! Wait...

This is a double-edged sword.

Management tracks are typically the next step up in terms of career advancement for regular workers. Eliminate those and what do you have? A couple mega-rich owners, and then a flat field of "peasant" laborers. It starts looking a lot more like a feudal fief.

What will take the place of middle management for career advancement?

Completely OT. What's with these articles that look like traditional magazines on the web? I get this is a PDF, but it makes reading on screen painful. I think something like the medium layout is the best way to present writings and you can still build fancy edits on top.

Yes, that really annoyed me about the article too. It made it much harder to read on a mobile device for one, plus I wanted to copy some text and share it as a quote with a link to the pdf but that was harder due to it being a pdf.

A Technology Freelancer's Guide to Starting a Worker Cooperative

http://techworker.coop/resources/technology-freelancers-guid...

A counter-example from HBR Magazine's December 2013 issue

http://hbr.org/2013/12/how-google-sold-its-engineers-on-mana...

cf. "Maverick" by Ricardo Semler, the CEO of Semco.

I think you'd all enjoy reading it. It changed my attitude to work and management considerably. I am still a manager, but I give a lot more responsibility to my team than most.

Interesting idea, but it seems to have a trade-off -- instead of negotiating a single (or perhaps a few) relationships, you have to negotiate about 20. It sounds a lot like a n*(n-1) problem.

The world is not a computer (or else AI would be easy).

For all of the companies who are successful without a management structure, their are exponentially more with a standard or modified structure that do quite well.

Any of those, "Let's get rid of management structure" companies on this list??

http://money.cnn.com/magazines/fortune/fortune500/2013/full_...

You might want to ask yourself why then. . .

I've come to a realization that managers and programmers have something in common. The bad ones try to perpetuate job security by creating complexity (whether in code or interpersonal issues) that only they can navigate. The good ones want to do their jobs so well that they "program" (literally for engineers, figuratively for managers) themselves out of a job, so they can graduate to better things.

You see this most strongly with consultants. The good engineer does the best work he can, assuming it will lead to more challenging and interesting projects in the future. He's not worried about job security, or at least not enough to do things that are unethical; he assumes that doing a good job and becoming better at his work is job security. The bad consultant obfuscates code, documents poorly, and tries to make it impossible to ever fire him. He's not trying to bigger, badder (and more lucrative) projects in the future; he's just aiming to keep whatever income stream he has in perpetuity.

I think that managers exhibit the same dynamic, and I think that solving this problem requires recognizing it and watching for the warning signs early on.

I'm strongly in favor of open allocation, but that's not quite the same thing as "no management", which I think might take the idea too far. Why? Because management is a fact of life; some people will have more power than others, and I'd rather it be dealt with in a fair and reasonable way than in an ad-hoc and unstable way.

Having a permanent class of entitled (literally, not necessarily pejoratively) managers may not be the solution, and I support making people more self-managing-- actually, I'd use the term self-executive-- but acknowledging the basic fact of management, and encouraging the positive manifestations while avoiding the negative, is probably healthy as well.

Honest question: have you ever met someone who purposefully obfuscates code? I've met some shitty programmers, but I don't think I've ever met someone doing it on purpose.

Purposefully? No, but I'll admit that during undergrad I mistakenly obfuscated code. We were competing in the Intelligent Ground Vehicle Competition (igvc.org), but I didn't really understand the "meat" of the problem very well. Instead of focusing on machine learning and simultaneous localization & mapping I got way too focused on the "multi-threaded architecture" of our robot and basically rewrote protobuf poorly rather than actually solving the problem.

I've improved in those areas since then, but I slap my forehead whenever I think back to those days about how stupid I was. The worst was that the whole time I thought I was so smart, too.

Boss has a problem, so he hires an MBA, now he has 300 problems, and he got fired as well.

Reply on news.ycombinator.com