Hacker News new | past | comments | ask | show | jobs | submit login
210 Zappos employees – 14% of the staff – take buyouts after CEO ultimatum (businessinsider.com)
425 points by jhonovich on May 8, 2015 | hide | past | favorite | 309 comments

I work at a company that was in bad shape during the financial crisis. A temporary CEO was put in charge. Ultimately, a number of companies were merged to create one company that could survive. In the end, one of the CEO's of the merged companies became the CEO of the final company. His first act as CEO was to offer everyone in the company a tenure-based package to leave the company. I don't know the full numbers, but nearly everyone in the IT department took the deal. And I am aware of many people in other departments who took the deal as well. Rumors were that between 30 and 40 percent of staff took the deal. Once this happened, the company got rid of that CEO and had to then offer alternative packages to people to stay. Basically the same amount you would have received if you left, but paid out to you over a few years to guarantee you'd stay.

It was unclear what his intentions were. To determine loyalty to the company, possibly. But what actually happened was that people who were good at their job, and could find work elsewhere, left. So they would have had all of the top talent leave the company and the remaining employees would be the people who did just enough to get by. The company is still trying to make amends with the employees.

I think the idea with Zappos is that they're supposed to be such a great place to work that you wouldn't want to leave.

And in Zappos' case, I get why they offered severance deals. They were embarking on a radical experiment in organizational structure, and not everyone is going to be on board with it. Rather than having a bunch of people at the water cooler talking shit about the changes, you offer them money to leave so that people who aren't on board with the changes get out of the way.

I would guess that many of the people at Zappos who took the deal were managers themselves. If you're a middle manager, you have skills around managing people but maybe not so much with doing the actual work. If your skill set doesn't align with the new job you're being asked to do (which is effectively a demotion), and the company offers you a buyout to leave, what would you do?

I imagine that people in management positions have reason to take the package even if their skill set is perfectly aligned to the new job. If you're on a career path where one of the major measures of your success is how many people are under you, then you better be pretty committed to the Zappos vision if you're staying on. A long stint of not having anyone reporting through you might be a bit of a black spot on your resume if you ever end up for a position elsewhere.

Yes, even the title change from something like "Senior Manager" to "Employee" might hurt your chances. With IT/dev it's not so bad, as some very good organizations list both grads and greybeards as "Member of Technical Staff." I have to imagine that those who remain plan to retire from there.

> If you're on a career path where one of the major measures of your success is how many people are under you,

My assumption would be that those are the people that this move is attempting to push out.

The list of people who took the offer was posted internally. I imagine it must be out in the wild somewhere by now as well.

Besides managers who are looking to jump ship and pretend to do work somewhere else, a lot of professionals are on the list- lawyers, accounting people. I imagine some of these people will show up on Amazon's roster in the near future.

> "Pretend to do work"

Little presumptuous isn't it?

I'm surprised the number of people taking the deal isn't higher. If I need religious zeal, I'll go join a religion.

well another way to look at it is that there is already a fundamentalist organizational religion that most people adhere to, they're just unaware of being part of one already

Please identify this "fundamentalist organizational religion" a bit clearer. Did you mean capitalism, b-school, emergent hierarchical organizations.. tribes?

its pretty simple, there is obviously a very ingrained faith in the standard organizational structure w/ management, staff. its so obvious that is how things should be (its "fundamental"), that its difficult to consider that belief is possibly also a matter of faith, even if that standard structure was itself emergent in some respects.

its possible the commenter i was responding to was referring more to the "rah rah" aspect of zappos as a whole, rather than religious zeal re: this new organizational concept. those are different things in my mind, also both are exhibited by the directive "to all employees: read this book and indicate you have done so".

anyway it isnt a comment on whether or not the zappos idea itself is useful and/or just another fad coated with new terminology (i have no idea). just that when you can identify someone else as engaging in religious zealotry, its fun to consider what that zealotry is being compared to, which probably turns out to be something like "everybody knows its supposed to be this way", aka. some sorta implicit religion.

Please explain.

There are reasons to stay besides religious zeal, too. I'm not certain how one can easily see this change as seriously threatening unless you're somewhere in the middle of the org chart. If you're at the bottom, hey, your day probably won't change that much (unless you hate your boss). If you're at the top, well, I've yet to hear that Mr. Hsieh is planning to abdicate.

Or perhaps you've experienced the Zappos culture first hand, and, for whatever reason, don't think that this is a good idea.

Three months is cool, but it's a big reordering of life. If you enjoy your job, you shouldn't take it. Maybe your next job is worse. And people in SV say "every good developer has 5 offers waiting for him by breakfast" but it's not that way in every part of the country.

On the other hand, if you've been contemplating a job change, take it.

I believe Zappos is in Las Vegas. And they're not really a tech company, although they use tech.

> they're not really a tech company

By your definition, what is a tech company?

My definition is that they use tech as their differentiator and/or as their value proposition.

You can be a tech company and sell commodities, Amazon is an example of that.

Amazon has not only used technology as their differentiator/equalizer, they've also innovated in other areas of technology and now resell that to others. Amazon is definitely a tech company on that measure. Zappos - they sell shoes. Now maybe they've done some innovating/inventing themselves, but I don't think they are known for that aspect of their company.

Not sure if you're aware, but Zappos is owned by Amazon.

Guessing he or she means they sell shoes, not software or hardware.

I'm surprised the number isn't higher because they recently moved Zappos from Henderson to downtown Las Vegas.

Recently? I'm pretty sure it was over 12mths ago as I toured the Vegas HQ in late May last year and it was well lived in by then.

I consider that recent. Long enough to know you hate the commute.

There's plenty of reasons to stick around that aren't religious zeal.

There certainly are but are there that many reasons that are good enough to make someone stick around instead of taking 3 months pay while not having to work?

I imagine most developers could find a job with similar pay in a month or maybe one and a half, so the reasons would have to be strong enough that you'd give up an extra 10% in salary, or even being able to take a 3 month long vacation.

On the other hand, if they recently had to move to Vegas I wonder how hard it is to find work there and I can totally see people not wanting to move again so quickly.

In the grand scheme of things, 3 months pay is not that significant (a large enough performance bonus can already match that, or a joining bonus for that matter). Now compare that against the chances that you will enjoy your new same-pay job less, or you'll have to move far away, or delay career advancement while you learn the ropes at the new job. There are plenty of potential advantages from changing jobs too, depending on the job, so it actually might make a lot of sense to switch. The point here is that 3 months of salary tips the balance only slightly.

Of course, those who have reservations about working where they are, should take the severance package, but that's the point.

Also, I am a bit surprised that we are discussing 3 months of pay as a magnanimous, almost foolish, concession from management, whereas in many places around the world that kind of compensation is law mandated when you lay-off people without reason.

3 months pay is nothing compared to years at a great company with great culture and great people. The work itself might be the same company to company, but the people and the environment depend a lot on internal culture that's not particularly easy to find before taking a job elsewhere. One spends at least 8 hours a day working with others, not just writing rote code. There's very much a human aspect to employment, not simply economic self-interest.

> I imagine most developers could find a job with similar pay in a month or maybe one and a half

Geographic area heavily influences this. Last time, I spent 8 weeks with almost no nibbles (and then suddenly had two companies fighting for me on the 9th week).

I've got mouths to feed. And when you interview for your new job, you are going to be asked why you took the buyout. "LOL 3 months salary" isn't a good answer.

Your point is valid. However, saying "LOL 3 months salary" would be a foolish answer in an interview even if you did take the buyout. Leaving a company over differences in philosophy (management, engineering, or ethical) are all perfectly acceptable reasons to part ways without compensation. I wouldn't take the buyout flatly, but if I wholeheartedly felt the direction the company was going in was foolish, I would certainly consider it.

3 months is less than 1% of lifetime earnings. Not worth throwingaway your plans (whatever your plans may be)

But factor in the value of 3 months. http://www.investopedia.com/terms/t/timevalueofmoney.asp

Not everyone likes the ping-pong ball model of modern employment. A stable job, where I show up five days a week, the company takes care of me, and I get to form stable working relationships with my coworkers, is valuable to me and many others.

Henderson is a suburb of Vegas. You wouldn't have to move unless you wanted to walk to work.

Some may still quit. They may just not have taken the last second offer. Instead, I wonder how many are interviewing and will leave once they have an offer safely secured. I'm sure Zappos will not want to publish that, because combining that will take them to probably 20+% turnover. Ouch.

Keep in mind that they're in vegas. There aren't a lot of similar companies there to transition to.

Rather than having a bunch of people at the water cooler talking shit about the changes, you offer them money to leave so that people who aren't on board with the changes get out of the way.

There's that element directly, but I think even more powerful is the sense to the employees that they are consciously opting in to staying. They are literally passing up a (small) pot of gold in order to stay in the new structure. The mere fact of choosing in such a way (even if that's the default choice) will tend to increase commitment in the inevitable dark moments of the transition.

Doesn't Zappos make a similar offer to everyone that finishes their new employee training program?

"Rather than having a bunch of people at the water cooler talking shit about the changes, you offer them money to leave so that people who aren't on board with the changes get out of the way."

Each change creates winners and losers. The losers will be around that water cooler. You can't manage this before the trajectory is flown. Best of luck to all involved.

it's easy to forget that zappos was acquired by amazon. i am interested to what extent sales growth or lack thereof, would be a driver of radical restructuring.

I think that Amazon is content to let Zappos experiment. If Zappos loses customers, statistically most of them will end up at Amazon anyway. Tony Hsieh is pretty out there, and a lot of his ideas sound kind of crazy, but I think Bezos believes in him and is willing to let him try. Even though everyone says "Oh, that won't work" you don't know until you try.

If Zappos can experiment with a new leadership style, maybe it produces desirable results that can be replicated elsewhere. Maybe it fails miserably. What's likely to happen is that there will be things that work and things that don't, and I guarantee Amazon is interested in that. Because of all the things I've heard about Amazon (many of them good), I've never heard that it's a great place to work. Their turnover is high, so maybe they're looking for ways to reduce it while not losing the culture of urgency (though some would say fear) that makes them so good at executing on things.

Of course, if it's a total failure, Bezos will likely fire Tony and trash the entire idea.

It's a great place to work. There, you can no longer say "you never heard it's a great place to work". Sample size of 1, but I love it.

Are you an AWS employee? They're the only people who ever say good things about working at Amazon in my time reading HN.

Nope not AWS

That's great to hear, but how much did that change for you or others when the holocracy processes were added? I can see where that could totally change how I feel about a job even if I still love my coworkers, the job perks, etc.

Sorry I should have clarified - I meant Amazon is.

That brings the total count of satisfied workers I have heard from to 2.

My spouse worked there, a couple of steps down from Jeff Bezos. At least parts of Amazon are micro-managed misery.

Amazon implemented pay ($2,000 -- $5,000) to quit about a year ago.

For warehouse associates. This doesn't apply to engineers.

This is the risk (the "wrong" people taking the buyout). It was the source of a lawsuit during one of HPs many acquisitions and then downsizes where HP tried to give "selected" individuals the option to leave early, and the folks who were not given the option sued. The courts affirmed that you could not give preferential treatment for early retirement/severance it was everyone had the same choice or no one did.

For what it is worth, managing staffing levels is one of the hardest things to do in a large organization because the visibility of contribution is so low. You have the quiet guy who gets things done everywhere but is invisible, and the useless guy who keeps up a very believable patter of what they are getting done. If you layoff people just based on the optics of the situation you screw yourself.

I think that's one of the important roles of a team leader - to know who is getting how much done (among many others). I wonder how Zappos will accomplish this now that they have no team leaders.

Yes, and another way to look at this is that team leaders maintain teams. They're not there merely for leadership's sake, they happen to have a particular role, maintenance.

Who will maintain teams without people whose specific role is maintenance?

I had a team leader who had to convince me that I was getting things done. One of the best.

Personally I think this idea is probably flawed in the first place.

a la 'Office Space'

People work at a company for a variety of reasons. Some want or need the money. Some really believe in the company's mission. Some like the people they'll be working with. Some want to develop skills for their next move. Some want the brand name or status. Over time, the company will accrete people of all sorts of different motivations.

What a buy-out does is take money out of the equation. It makes it so that everyone who is there for financial reasons has a financial reason to leave. This can be logical if the company has little to offer financially, as in, for example, a company that's in bad shape during the financial crisis. The only employees who're left are those who are not in it for financial reasons - the true believers in the company, those who like their coworkers, and those who want to gain skills. Those people will be a lot happier when they don't have to listen about their financially-motivated colleagues grousing about how the company is going under, and they'll be more productive when they're happier.

The CEO's fatal mistake was that he forgot that his board is financially motivated. From their perspective, he's acting insane, because why would anyone stay at a company for reasons other than money? And they're his bosses, so they can have him removed. For this to have worked, he probably should've taken the company private, or at least packed the board with people who bought into his vision for what kind of company it should be.

The board's fatal mistake was in removing the CEO. By offering buyouts, he had just aligned the company in one direction. By removing him, they suddenly took the company in a different direction - one that they had less than zero capital in, because the CEO's previous actions had alienated everyone who might've helped them.

In both cases, the failure was in understanding that people are different and may have different motivations, and that a company survives based on how well it can hold together people who all have their own goals and worldviews and motivations.

I don't see how the usual buyout amount would be able to take the money out of the equation. "Low performers" who don't have good chances in the job market surely will stay, since their earning potential is greater than what you reasonably offer them in a buyout. The likelihood that the skilled employees will take the money, knowing they will get another job immediately, is much higher.

The buyout is so that you get mediocre, financially motivated people to leave. The idea is that you give them enough cash that they can be reasonably sure they will find another job before running out, but not as much as you would pay out in salary if they stayed.

Imagine your employee base graphed on two orthogonal dimensions: their performance level, and their motivation for working for you. (In reality, "motivation for working for you" is actually many dimensions, and performance is likely to be somewhat correlated with certain motivations, but this illustrates the point at least.) The buyout offer doesn't affect people who aren't financially motivated, so take them out of the equation, and look only at the folks who are strictly financially rational.

You can loosely group them into high-performers (those who are known achievers, who always have an offer in hand as soon as they make overtures toward searching), mid-performers (those who pretty much amble along, and for whom the job search provokes intense anxiety, because they typically have to put in 3-6 months of full-time searching to find a new position), and low-performers (those who don't do anything and everyone already knows it). In most companies, the distribution of these is roughly 10-80-10. You can fire all the low performers; the point of bucketing them into the "low performer" category is that their underperformance is obvious enough that they leave a documented paper trail that makes this unambiguous. And in a company that's run into financial difficulties, the top 10% has probably already left, because they continually get job offers and when your company is going down the tubes, it's to their advantage to take them. That means that most of your financially-motivated employee base probably fits into the middle range, the folks who are average employees and will have average difficulty finding a new job.

The point of a buy-out is to reduce the anxiety levels surrounding getting that new job to the point where anyone who is unhappy at your company but isn't sure they have enough money or skills to go elsewhere ends up taking the leap.

"What a buy-out does is take money out of the equation. It makes it so that everyone who is there for financial reasons has a financial reason to leave."

Unless you're not confident enough that you'd be able to get a job in those X months, either because of your skills or there not being jobs around.

Also, even if you like your coworkers, or want to acquire skills, if the ship is going down and you feel confident enough that you'd be getting another job quickly, then you might jump ship anyway.

The company I worked at "solved" this problem by offering the same package but with a clausing stating "key employees are not eligible to receive this package". There was some legalese at the end of the document that deciding who was "key" was ultimately up the company and no explanation was required.

As one of my co-workers explained it: If you're incompetent you can get a great payout and then go be someone else's problem.

That sucks. If you ask, you are signaling you want to leave, which you don't want to do unless you really can.

Isn't that illegal, though?

Possibly, but I have no idea really. You'd have to sue them anyways to find out wouldn't you?

> It was unclear what his intentions were. To determine loyalty to the company, possibly.

Corporate loyalty has been dead for many, many years.

Corporate loyalty died when companies stopped paying pensions, or severely reducing pensions. Our parents always worried about pensions and corporate loyalty. In our generation, and especially technology, we make a lot more money, and can generally make even more by switching jobs. Therefor, corporate loyalty, in the current tech industry, is virtually dead.

The one case that it isn't is perhaps companies like google where an employee gets the founders award, which is millions of $$$ of stock options that vest over a relatively large stretch of years. I know a guy who won this award and he is google for life.

Corporate loyalty is a two-way street.

In medieval times, the vassal paid homage and swore fealty in a commendation ceremony, and the lord granted him a fief. The vassal had a duty to aid his lord, and the lord had a duty to protect his vassal. The vassal gave loyalty, and got a secure income in return. It was a permanent relationship. To dissolve it, you had to use an axe, rather than a pink slip.

Corporate loyalty died when management culture forgot that the company owes duties to its employees in addition to its shareholders. Defined benefit pensions became defined contribution retirement plans. Then the employer contribution disappeared. Health insurance plans dwindled from all medical expenses paid to you paying 80% of the premium, $30 co-pays at every turn, and deductibles that are met midway through the twelfth month of the plan fiscal year, then reset. Training programs evaporated. Severance packages got smaller, or vanished. Even the amount of notice I get whenever permanent layoffs are to occur has become days, rather than months.

So there are no more loyal vassals. We are all mercenaries now. I'll start sending out resumes if I so much as notice that the building has a new janitor.

  H. R. tells new hires:
  "How great it is to work here!"
  Give me cash, not lies.
If corporate loyalty is dead, it is because corporations have killed it.

Exactly. There's no "there" there.

The death of pensions really did it. My wife and I each stumbled upon positions with a pension, and it seems as if the roller-coaster just stopped. We had no "loyalty" and moved around whenever the mood struck us, but now we look at the risk/reward of a more exciting job vs 25 years of stability and a secure retirement... it's not a hard decision to make.

Just be careful. I used to work at Nortel in Canada which at one point had over 90k employees. When I worked there, I believe they were making $18 billion a year in revenue. Profits appeared to be good.

A few years later it turned out the "good profits" were actually accounting fraud. The company tanked and everyone lost their job. But Nortel had a fairly good pension system, so the people who had worked there for 25 years or more were fine, right?

Well, the company feared that while they were tanking the "best and brightest" executives would leave. Oh my! What to do? They were allowed to siphon the pension money into "executive retention bonuses". Of course this didn't stop the company from ultimately running out of money, but it did get rid of a large part of the pension money.

Investigate your pension scheme and make sure you understand what will happen to your money should the company go out of business. Especially find out where the pension system is ranked as a creditor. No company is too big to fail.

As for me, I only worked there for 5 years and left when it was at its height. After I quit I got a letter from Nortel stating that my minimal benefits were not worth it for them to administer. They were going to keep my pension money and I was welcome to sue if I didn't agree. I guess as it turned out, it didn't make much difference...

Listen to this guy. I'd much rather get a big 401k match and have the money be mine than trust that my pension will be there based on the whims of the company and market.

Look at the airline industry. Plenty of pensions have been restructured during the string of bankruptcies. Pensions are guaranteed by a federal body, but only up to ~$50K/yr I believe. You'll never get less than that, but you $100K/yr pension might get cut in half and the gov't will say "that's the best we can do".

Yes, these days a private pension isn't worth the paper it was written on in many cases.

These things should always have been administered by 3rd parties and required to be fully-funded.

No "executive" can look at so much money "just lying around doing nothing" without wanting to get their grubby little hands on it.

Very sad for a lot of people

It wasn't mentioned if the job was public or private sector. Many government jobs have good pensions.

In Australia, this gets taken out of the equation as we have compulsory superannuation.

Could not agree more, the only downside is that some companies see you as a job hopper, but you can just explain to them that losing 30K-40K/year (aka staying in your job) is not an option

Is 30k-40k really correct? That seems a bit high to me.. If you could make even 30k more a year from switching jobs, and you started at the standard 100k, after 10 years you would be making 400k.

I haven't heard of any developers who make 400k. Then again, I suppose you never said this was for developers, so is there another job where this actually applies?

There's only been one documented (via a lawsuit) in the 400K range, although that was a few years ago: http://adtmag.com/articles/2011/05/27/what-highest-paid-prog...

I was following this pattern: test engineer, administrator, junior developer, developer, senior developer, principal developer, architect, principal architect.

My last 2 job changes were in the 20 - 40K increase range.

250K is not unheard of. I guess this caps out somewhere, but if the US immigration keeps H1Bs under this control than we are going to soon experiencing 300K wages.

Ah okay I think when you said job change I just assumed it meant going to another company, not having a new position or title. I do wonder how quickly you can change positions though. I imagine once you get to senior it's very difficult keep going higher. At least this was what I heard at Amazon.

Do you know if it's easier to get promoted when switching jobs? I have a hard time believing that since I think you would have had to accomplish something significant on several projects to get to principal, but if you switch companies it's very hard for them to truly the impact of what you did.

I think you are overestimating companies... For most of my career (12 companies, 3 of them my own) it has been easier to get hired at a higher position in a different place than being promoted internally.

When you are applying for a new, higher, role having enough experience at your current level, being sure enough of yourself and demonstrating the ability to create value within the test period will be more than enough for most companies out there.

When trying to get promoted internally you have to deal mostly with office politics. In a way, the quality of your work is only a small part. And, if you are good enough, it can go against you, I know people that have been skipped for promotions for years because their managers knew that the whole department depended on them.

Of course, there are exceptions and I admire those companies, but they are just that: Exceptions.

Interesting. Well, you would know better than I do since you obviously have way more experience! I appreciate the advice and will keep that in mind!

Staying with one company is hard. Especially getting promoted in Amazon, that is crazy crazy hard. Your best promotions are when you are changing jobs, I think the best time is to change when you can get a more senior position with obviously more responsibilities but also more recognition.

In SF this can be achieved quick, enough if you demonstrate your knowledge during the interview loops. For certain positions they also ask for previous experience but those positions more of the management side I think. If somebody walks in and writes good code we are willing to hire even into senior position, I don't care too much about the history. I have seen a candidate walking in to the interview from a grocery story where he worked as a cashier. Not only he got hired based on the interviews but he became one of the top performers on the team.

I think it is definitely easier to get "promoted" by changing jobs, especially if you are in a meritocratic environment, like some of the startups in SF.

A 10 or 15k pay increase per job hop should be easily achievable by much of the Hackernews demographic/stereotype. Sticking around several years for a single digit percent annual payrise could easily cost you 30k/yr over 2 or 3 years compared to switching jobs yearly or faster.

The speed with while a typical ~100k dev role gets promoted or awarded payrises to become a 200k position is _much_ slower that the dev who cherry-picks better opportunities when that arise, and works out close-enough-to-truthful explanations for any resulting 6 or 12 months roles (usually with ex-colleagues who'll give references backing up your post-role-claim that it was a 6 or 12 month contract role, not a permanent position.)

You absolutely can make 400k or more by working as a developer in finance. I work in finance and have made more than that some years at my previous job.

Hey I appreciate the input. I'm actually EXTREMELY interested in finance so I'd love to know more about it.

Do developers in finance require any unique skills? Also is there anything special one would have to do to break into the finance industry?

I guess analytical skills, math probably, but these are pretty general. What unique skills are you thinking about?

I don't have the slightest idea since I don't work there. I was more wondering if there was anything I would need to know before applying specific to the industry or if it's just like any other software engineering job.

To succeed in financial dev you need to be comfortable working in a heavyweight, slow large-corporate environment. Think lots of meetings, lots of time estimation & metrics, using Java, C# or C++ (shudder), people caring what you wear. If you can deal with this, the rewards are great. If you can't, it's hell.

There are finance startups where the above doesn't apply, but they don't pay like the banks do.

What you describe in the first paragraph aren't roles that pay 400k...

Math... and a lot of it. Look up job descriptions for a Quantitative Developer or Quantitative Analyst.

I got a $40k raise when I switched jobs the first time, and I was paid pretty well before then (not quite $100k, but I'm not in a tech hub either).

Then again, that's the only time I've ever done it, so I can't really say at all whether that trend might continue (almost certainly not).

Plenty of developers at Apple, Facebook and Google making 400K per year total comp. They aren't the majority, but there are enough of them to not be rare.

I've worked at companies at a minimum of 2 years or more and have switched between 5 companies. Would that be considered job hopping?

If I was a hiring manager, seems totally reasonable to me.

I have been at 6 companies in 10 years and have not had a problem explaining it. In tech, 1 year per company is reasonable as a minimum.

Same here, I had to explain to a HR manager once, but I told them if they don't get it they are not looking for me. :)

Maybe at Silicon Valley startups but it will raise flags in many places. That may be fine based on your interests however.

Then they'll simply complain that you're only "motivated by money" and that you "don't really want to be part of this team". This is, of course, ignoring the fact that everyone is motivated by money, and almost none of their employees would still be working there if they stopped paying.

> Corporate loyalty has been dead for many, many years.

The higher-ups in be corporations have yet to get that memo.

They got the memo; they just want to have their cake and eat it, too. They want you to be loyal to them, but to retain the option to dump you on the street as soon as somebody needs to bump their profit numbers by 0.1%.

Since we've learned that they are almost all sociopaths (not as an epithet, but clinically so), this behavior is perfectly consistent.

You missed a zero or two on the right side of your decimal point.

Surely you meant "left side".

Maybe you need to write a 5000-word memo to get the message across.


Corporate loyalty is gone, but nepotism never dies.

Loyalty runs both ways. Employees aren't going to want to stay loyal to a company in this mess of a capitalist economy.

Not for lots of people. And startup loyalty is even worse.

> But what actually happened was that people who were good at their job, and could find work elsewhere, left. So they would have had all of the top talent leave the company and the remaining employees would be the people who did just enough to get by.

This reminds me quite a bit of a blog post from The Wetware Crisis on what he deemed The Dead Sea Effect ( http://brucefwebster.com/2008/04/11/the-wetware-crisis-the-d... )

While not a forced evaporation which you describe, its the 'the best people can get jobs elsewhere' and 'the less capable people entrench themselves as experts on critical systems to where they can't afford to be let go'.

> It was unclear what his intentions were.

When you don't know how to run a business, cut costs and call the result a profit. Then take the bonus for "hitting the target", and pull the golden parachute ripcord.

This. I was going to post a reply just to say that, but I couldn't have said it better..

I once started a gig at a broadcasting company that had just bought a respected multimedia/internet company running several major high traffic websites. The broadcaster didn't like the employment contracts with all the "cushy" benefits, so they offered the staff a buyout. The entire technical staff took the buyout, most of the easily replaceable editorial, project management and graphics people stayed put.

I've seen this done, but there was a class of employee not eligible for the buyout (e.g., engineer).

So you wind up with good engineers, and all the bad management and support that couldn't find jobs elsewhere, and that works out real well, you betcha...

>But what actually happened was that people who were good at their job, and could find work elsewhere, left. So they would have had all of the top talent leave the company and the remaining employees would be the people who did just enough to get by.

I remember this being an issue with General Motors near bankruptcy (2008) too. They couldn't lay off union workers, and figured it would be cheaper to offer them buyouts than keep paying their above-market salaries.

There too, people pointed out that the only takers would be the best workers, who can easily find work elsewhere, and they'd be left with the ones who can't.

It was the same at British telecom when there where such good incentives to leave that many of the key technical staff left.

Most of the Uk's mobile networks was designed and built by ex bt people who took redundancy

>It was unclear what his intentions were.

"a tenure-based package to leave the company"

People who have been longer with company are also more likely to be older than average employee. So I think he wanted to get rid of "old" people who would "slow down" company and be unfit for its "new and vigorous" culture?

In my (thankfully limited) experience, it's entirely standard for severance packages to be tenure-based and, IMO, most of those reasons have nothing to do with ageism.

The longer you've been with us, the more you've likely contributed and we should recognize that as we part ways. The longer you've worked with us, the more life upheaval the separation represents. The longer you've worked with us, the more personal relationships you have developed, perhaps with the same people making the decision to eliminate your position. A generous severance (which IMO requires a tenure component) sends an important signal of stability to the employees who remain AND makes it easier for managers (who are of course humans themself) to make the decision to eliminate the position of a long-tenured but "surplus to ongoing operations" employee.

What's the alternative? The person who joined 6 weeks ago gets the same package as the 10-year veteran? That seems worse all around to me.

There is a subtle difference between company wide push to get rid of employees and severing employment on an individual basis. Giving presumable older people more in the first situation looks like a systematic effort to get rid of them.

Did he make the offer without consulting HR? I've received occasional buyout offers - they're always subject to approval and capped based on a number of factors.

I'm not sure how to gauge this. I think Holacracy may be one of the dumber pseudo-science management theories that's ever come out. If it works for Zappos I'll eat my hat I guess, but I have a feeling that all we'll hear about it is that "it's working great!" as the ship goes down in flames.

Holacracy and other ultra-flat org structures always share two features:

- It consolidates power in the people at the top (the people who do not participate in the flatness of the new structure)

- as social animal, humans will always naturally self-organize into de-facto hierarchies that may or may not be aligned with the needs of the organization

Prediction, the 14% that took the buyout are probably the ones that couldn't fit into the new clique driven de-facto structure and were kissing years of promotable work goodbye. These folks will likely be successful wherever they go. Meaning that the people left behind have either cemented a position in the new structure through force of personality (not competency) or are not otherwise employable elsewhere and will suffer as the assholes fight it out on the bridge.

Here's an employee review


Seems to work for Valve, with a few obvious problems: http://en.wikipedia.org/wiki/Valve_Corporation#Organizationa...

Valve is the exception that proves the rule, i.e. you need a stable cash cow so large that your frequently-aimless often-delayed progress on anything else doesn't break the company.

Here's Valve's reviews


Here's some choice quotes:

"While it's true that Valve has no official job titles or promotions, compensation varies greatly among employees and many teams have an obvious pecking order. There is no formal management structure, but it's clear that some people have substantially more control over project direction and the work of others."

"Because teams are intended to be self-forming, it's rare that enough people will want to assume risk to all collectively embark on a new project. It's too safe and too profitable to just contribute to something that's already successful. Even though failure is supposed to be tolerated and even encouraged so that employees will try new ideas and experiments, there is little evidence of this. After a few rounds of bonuses, folks learn quickly what is rewarded, and what is not."

"The idealistic paradise is ultimately undone by a flawed review system. The lack of managers means that a peer review system is necessary, and Valve is very proud of theirs. But their review model is best described as a "popularity contest masquerading as data""

"Those who get stock options do extremely well, and the others do not. It’s an unacknowledged two-class system."

"The rational response to this uncertainty is to find a patron – somebody who can guarantee you a good review if you do their bidding. These patrons (the knights) guarantee themselves good reviews by doing the bidding of a higher-level patron (a baron), and the barons pledge fealty to the board members. This unofficial structure necessarily evolved and you opt out of it at your peril. The irony of a hierarchical structure spontaneously forming in Gabe Newell’s company after he has spoken so strongly about the problems of “command-and-control type hierarchical systems” is delicious. As was noted in “The Tyranny of Structurelessness”, “structurelessness becomes a way of masking power”, and this masked power is more insidious than formal power."

"So, I quit in order to get better compensation, an acknowledged hierarchy, and appreciation for my work."

"This organization has a purposely opaque, hierarchical, secretive, and very rigid management structure."

"Some projects can go on literally for 5+ years wandering around pointlessly without shipping, with little to no direction, and no accountability. This company is terrible at writing and shipping large scale software"

"The random mass firings of 2013 tanked moral, and the stream of talent leaving the company during 2014 didn't help."

"The yearly review process lacks feedback, transparency, and coverage. This company has no formal HR, so good luck if you need to give genuine feedback about troublesome coworkers."

"The lack of solid structure in management reduces the company to schoolyard politics where bullies and loudmouths reign." (Title: "It's like being in high school again")

"The culture at Valve is a bit like a cult. There's a party line and if you veer from that, it is discouraged with one-liners rather than discussion."

"The incentives are setup so the people who place themselves around upper management the most and are the loudest about what they're doing (that jives with what upper management likes) will be compensated several times more than those who don't."

It's interesting to note that these come from various ex but also current employees. Talks a lot about the culture over there.

> "The random mass firings of 2013 tanked moral, and the stream of talent leaving the company during 2014 didn't help."

I think this one is talking about the departure of Michael Abrash (and several others) to Occulus.

And pretty much all of the kinds of obvious things that you'd expect to see in a "self-organizing" environment are called out.

- de-facto hierarchy takes its place (at valve it seems that there are several overlapping hierarchies, all of which are toxic)

- highly centralized power at the top results in the de-facto highest people jockying for favor and acting as gatekeepers

- innovation stops

- complex projects take forever

- risk taking gets turned into position jostling

- unclear structure leads to unclear reviews

- nobody to resolve disputes

- secrets and rumors dominate information flows

- these problems remain unacknowledged

Look at reviews for other large flat organizations and you'll see the same issues echoed over and over again. Anybody who's worked for a reasonably sized "flat" org will recognize all of these problems as persistent and near universal.

Zappos will be no different. It's just a terrible way to organize labor.

Talk to employees and you may hear differently.

I've heard the same. The answer to bad managers and overly complex corporate structure isn't to completely get rid of all structure and managers.

That review sounds exactly like what I expected. It's more a rah-rah you're with us or against us mentality. Sounds rather cultish even.

It's like places saying they're okay with flex time but you get poor reviews by butts-in-the-seat managers regardless to your actual performance.

This is why I'm simply riding the clock at my current job. It's sucked the life out of me. Change is on its way fortunately.

That's my prediction as well. Without a designated leadership structure, it will devolve into high-school-esque in- and out-crowds where the cool kids and big-talkers run the show and the smart kids and do-ers leave.

Motorola pioneered six-sigma, i guess Zappos wants to pioneer a brand of management style (sans management).


This is probably the most fascinating response I've ever received by a non-newbie on HN.

I'm not even sure how to think about it. It's like some kind of weird attempt to get shadowbanned or something.

In this case "I'll eat my shoe" might be more appropriate.

"Avoid gratuitous negativity."

I would categorize that as humour, actually.

Thought experiment: ignore the holacracy angle and offer everyone in any company 3-months severance if they quit by a certain date. What percentage would quit?

I wouldn't be surprised if the average was 10-20% and a vast majority of these people are taking advantage of what they see as a good deal irregardless of their feelings on holacracy as a management style.

Actually, that aspect of Zappos's strategy isn't new at all.

Here's an article from 2008.

tl;dr is "If you quit today, we will pay you for the amount of time you’ve worked, plus we will offer you a $1,000 bonus."


"tl;dr is "If you quit today, we will pay you for the amount of time you’ve worked, plus we will offer you a $1,000 bonus.""

Paying you for the amount of time you worked is a legal obligation. So really you are talking about can you find an equivalent or better job in the amount of time you would have earned $1,000 if you stayed. Even in the lowest of paying legal jobs in the US, it does not take that long to earn $1,000 relative to securing a job. Not to mention that new job is not guaranteed to fit into your current living situation and other factors that cause the inertia for staying with a current job to be quite large. Seems a bit like a gimmick to me.

True, though 3 months pay for seasoned employees is signifanctly different in timing and scale

What percentage would quit? The ones that could land another job by the quit date. They would essentially get a 3-month "bonus" for doing nothing. I would say this is only possible if the these people were looking to jump ship anyway.

That is assuming that everyone is a hired mercenary. I could rather easily find a job that pays me a 10%-30% raise over what I currently make. I don't do that because I already make "enough" and I like my current job, company, and coworkers. It isn't even a matter of loyalty, it is a matter of different priorities that puts money further down the list than job satisfaction. Finding a good paying job is much easier than finding a good job.

I can find a good job easier than I can find $30k+ free money.

If you can find a better job than you have, why haven't you done that already?

No one's giving me 30k to go look for one.

This is highly relevant. Looking for a new job is like another job in itself. You have to do research, have telephone conversations, be "switched on" in your interview persona for hours at a time, jump through flaming hoops, and if possible, also be younger, more energetic, and more beautiful than you really are.

It's a sales job, with you as the product.

I. Hate. Doing it.

So while a lot of people could go out and get a better, higher-paying job, there is a potential barrier and significant activation energy for that reaction. Doing my regular job plus the self-sales job is absolutely exhausting.

When my current job is already almost good enough to keep, I have little incentive to go sift through piles of recruiter spam and speculative postings to find the actual jobs, retype every last detail of my resume into yet another stupid custom candidate tracker web-app, suffer through the bozo filter and the pop quiz of the tech screen, fly or drive wherever, and spend one of my vacation days playing along with some stupidly involved interview process, only to see a curtain of silence descend forever.

For $30k, and an extra 40 hours per week in which to do it, I would certainly put up with the tech hiring circus for a couple of months.

Note that he said a good job, not necessarily a better one. I think the idea is that if someone is giving you 30k to find one, you really only need to find one that pays the same, or close.

Technically you'd come out positive if you found something that's at least more than (what you make) - 30k, but then you have to consider raises and such.

There's always a better job. At some point you have to stay somewhere for awhile, to gain whatever it is you're trying to get: stability, experience, improvement, contribution, pay, benefits, community.

Their current job might be "good enough" in that the benefits of switching are not significantly over the costs of switching. But when you add 3 months of salary to the equation, that might tip the scales.

hear hear

There is nothing wrong with being a hired mercenary.

If you consciously attempt to hire true believers, then you will get two kinds of employees: true believers, and people who lie about true believers.

That set of employees is probably pretty small compared to the larger set of qualified employees without such a restriction. You write off a huge pool of talent.

Just because you're a mercenary doesn't mean you won't bring your all to the task at hand.

Also, consider the employee relationship from the other direction. No corporation is going to hire an employee because the corporation is a true believer in the employee's mission. That relationship is a strictly mercenary one: Can you do for us what we need you to do for us?

Full disclosure: I'm old.

But if you weren't currently happy with your job and you weren't confident in the chances of that changing, it wouldn't make you a mercenary to jump ship at that point.

When it comes down to it, it's a matter of comfort and risk.

What percentage of any competent IT team can't have a competitive offer in writing by the end of the day?

Even if your anticipated time between jobs is 1-2 months, you can still confidently quit by the deadline and end up with six weeks paid vacation and a six weeks bonus.

In the US I would guess around 95%.

Extend things to 1-2 months and that probably drops to 20% or so which is very good odds.

However, the difference between a 'competitive' offer and a 'good' one at a company you want to work for is huge. If your spending between 1/3 and 2/3 of your waking hours at work you need to make them count.

Furthermore, Zappos is in Las Vegas. It wouldn't be my idea of a place to live but some people like it and it's certainly hugely cheaper than most of California. So someone leaving Zappos in IT/software development probably has the option of going to work for one of the casino groups or moving.

A job offer by the end of the day? Seriously? Or did I misunderstand you?

You still have to schedule interviews, take vacation to interview because it takes all day, etc. You should probably interview at many places to get a better offer, and because you can get rejected for pretty silly reasons. You might want to review your whiteboard coding too. It can take a few weeks / months.

When you refer to IT do you mean IT/desktop support or software engineers?

Not everywhere in the US. Some places are full of mostly crap jobs.

How many people get a three month, paid vacation at any point in their life?

It's not a vacation if you don't have a job to come back to. I'd relax that first week, but then I would need to start job hunting. And this thought that "you can always find a new job" quickly might turn out to stop being true tomorrow, or not be true for your exact specialty or geographic location. (Especially if 210 other employees hit the job market the same time as you.)

Letting fear rule your life makes for no fun.

Intel still has two month sabbaticals AFAIK but, yeah, it's incredibly rare.

It's not even so much about the vacation being paid but having certainty. I haven't changed jobs a lot but the last time was pretty typical: get an offer and then "when can you start?" I ended up taking a 2-3 week vacation arranged on about 3 weeks notice.

> The ones that could land another job by the quit date.

The ones that are confident they could land another better job by the quit date.

It's not enough to be possible to get any job. The offer only makes sense if you are certain you will trade up.

I doubt those 14% are that confident. Confident makes it sound like it is a sure thing. I think it is more like: If they could, they would.

And it does not have to be a better job, it could be even one that pays less. If the person needed cash fast, this is one way of getting it. They could spread those losses over a year. It's still better than a payday loan.

Why better as opposed to one that is just as good? If you got another job that's just as good that'd be akin to getting a 3 months paid vacation. I would LOVE that

Or the ones that think they can find another job with 90 days, or the ones who want to try their own thing for a little while (subsidizing the first 90 days of startup runway: enough for a MVP).

And not happy with current job.

If I'm happy, I'm not quitting even though I could easily get another job. It's not always about the money.

One important subtlety here is that it would depend greatly on where you were geographically. In a small town with one employer it would be one way, in LA or NY it would be another.

Zappos is located in Vegas, which is a bit of an oddball market and somewhere between these poles I would think.

I think that only people who do not have particularly good relationship with the rest of their team, and either already have or can easily find another good job. If that's a somewhat stable company, I don't think the percentage would be high. Most people don't like change. If things are okay-ish and you have friends on your team, I don't think a 3-month severance is enough to make you change your comfortable routine. Do some large change at the company, like changing the management structure, moving the company and the chances that people will take that offer will very significantly increase.

It's like cleaning your email subscriber list every few months.

Not really; people who can't get a job elsewhere possibly because of skill wouldn't be taking the offer. But someone who gets constant job offers by other companies? Unless you're absolutely in love with Zappos why wouldn't you? I know I would have.

Fair enough. Though I'd note that there's quite a few options in between "no job prospects / low skills" and "constantly getting job offers". The average complacent / comfortable employee probably does more harm to the company than most others on a daily basis. I know PLENTY of workers who stay in jobs for years while being unhappy and constantly talking of wanting to leave. That's more of the "list cleaning" to which I was referring.

Although of 1500 Zappos' employees, I doubt many are getting frequent offers or recruiters reaching out. How many are doing IT, development or high level sales stuff where well paid jobs are plentiful? I imagine the majority are customer support agents of one kind or another, packers, logistics people, etc.

This sounds a lot like the flat structure they have at Valve, except at Valve they have had that since Day 1 and everyone is used to it. I can't imagine the headaches and the absolute confusion that would result from trying to take a traditional management tree and flatten it. It seems like by its very nature the kinds of people you hired when you had tiered management might not be able to easily transition to a flat management, nor would they want to. Having worked in both environments, I can see that there's advantages and disadvantages to each, but I personally prefer having a manager to bounce things off of and to (for lack of a better term) shield me from any unpleasant office politics.

I suppose I've just had good managers when I've had them, which is of course going to make things work out well for me.

There are articles on how Valve has "hidden management" cliques though [1]. I'm not sure if this was ever addressed by Valve itself.

[1] http://www.develop-online.net/news/valve-s-perfect-hiring-hi...

It's very similar in holocracy: the people who are natural leaders will fill the management needs of the organization; it's just more chaotic, less predictable, and not as good. The lack of structure brings out the worst in human behavior.

In my experience, it is a horrible system in which to exist. It's highly ignorant of psychology and behavior theory almost intentionally, and it creates far more drama and politics than any organizational system I've ever seen. It's not a good thing.

Yeah. It's essentially creating a power vacuum.

People don't work without structure, ad hoc structures develop. The conjecture is that this is a good thing, that these ad hoc structures are more natural, and you will have natural leaders take lead where appropriate and this will be more effective.

But the issue is structure deals with quirks in human behavior. It deals with the emotional side of it, it deals with the cliquish, tribal nature of our default social behavior. You can put a highly productive, less social person into a position, give him a set of tasks to do, evaluate him fairly and see him produce. In doing this, you need an objective manager, you need a fair structure, and it works.

Without those things, the less social, highly productive person is easily sidelined. The person who takes leadership isn't given that position, they don't lead because they have been shown that they can motivate people to do great things, because nobody gives them that position. The people who get into leadership positions are those people who can convince other people to work with them on a personal level instead of on an organizational level. The really smart, less social person doesn't even get brought in to participate, and little power struggles happen between factions of people who want to be the person who directs the project, despite the fact that none of the people looking to do it might even be very good at doing it at all.

The real problem with traditional management structures is that it often breaks down to this sort of system to start with. Where the people building the systems aren't very good at it, and where they invite their buddies to be management because management pays the best, and they want to help their buddies out. It's because you have poor systems, and because you have the wrong people managing.

The issue with this is that rather than potentially falling short and missing the mark, they start with the worst of it.

> The real problem with traditional management structures is that it often breaks down to this sort of system to start with.

This is exactly right. The traditional organization already falls prey to politics and games due to lack of good structure, not due to the idea of structure itself.

You will get a system regardless of whether you intentionally create one or not. It's far better to get the one you intend and can develop, than an uncontrollable chaos.

> The lack of structure brings out the worst in human behavior.

Well, that's your experience, and you're entitled to it. I'm finding the opposite. Though describing holacracy as a "lack of structure" seems like a pretty serious mis-statement. Holacracy is extremely structured - way more, and way more explicitly, than most organisations.

It's kind of like the classic traditional observation that "agile is basically no processes". Yeah, if you use agile as an excuse to throw out process, you might equate the two. But actually agile (at least at its beginnings) involved a lot more process and rigour than traditional waterfall development. Agile is not an absence of process. Holacracy is not an absence of structure.

I guess humans will be humans - which means they will misinterpret, distort, exaggerate, etc, especially when exposed to newness.

I have little comment about what holacracy is like to work in, because I haven't worked in one. But I started a business that now operates by principles fairly closely related to those underlying holacracy (though the implementation is fairly different). We have less structure than holacracy, however... and yet what I'm seeing is that quite the opposite of your statement, it is bringing out the best in people.

Then again, I've not implemented a fake holacracy for all the wrong reasons. We're still in the relatively early days of this sort of culture. There are many more ways to get this wrong than right. Unless the people making this happen are truly committed to creating an open culture, and are doing it for the right reasons (i.e. this is the way they want the business to operate because it's good - not "well, this seems like a good tool to get people to be motivated without paying them nyeh nyeh nyeh dr evil laughter"... unless it's done right, it will probably be a horrible distortion of what it's meant to be. Kind of like what you describe.

I think you should consider the possibility that the allegedly holacratic organisation you worked in just got it really badly wrong.

I think that's very true, and I appreciate your perspective. I have no doubt it's very truthful, and there are some great things about the openness that might be possible with a holocracy or something like it.

I agree that the organization I worked in got it really badly wrong, but I also saw how the structure that was being followed (and there is a structure) sort of enabled that. It was a small organization and there were many other things going on that I think caused issues—so perhaps it's not 100% attributable to holocracy, and wasn't perhaps the best situation for it to be effective because of that.

I'll admit it was anecdotal. But I don't trust a process that, when followed fairly closely, had what appeared to be many holes in its organizational model that were suboptimal.

Are you speaking from experience? It seems that if you completely do away with job titles, the politics would go away right?

Politics isn't tied to job titles. There still are people with different amounts of influence that try to influence others for their own goals. FWIW, Jeri Ellsworth (who was hired to do hardware stuff, didn't do well there and was fired in the end) described office politics at Valve "like back in high school". Not saying that it can't work, but it also is not without its flaws.

I wasn't aware of who Jeri Ellsworth was, but a cursory look says that she was fired by Valve.

Were her comments before or after she was fired? I'm not sure that if it was after, we can necessarily count on it being objective. I've heard plenty of people make comments that aren't exactly the most accurate after they've been fired.

It's very difficult to deduce causality in that situation. She could have been fired due to not fitting in with the organization due to her valid concerns, or she could have brought up concerns due to bias from being fired. It is most likely somewhere in the middle, where her observations were somewhat truthful and somewhat biased.

http://www.develop-online.net/news/valve-s-perfect-hiring-hi... , which is one of the parent posts, gives some of the timing details.

Even if it were before, would you "necessarily count on it being objective"? It seems like an impossibly high standard, given how many people make comments that aren't exactly accurate before they've been fired.

> given how many people make comments that aren't exactly accurate before they've been fired

That's certainly fair. I'm not making a comment one way or the other - I don't know her, I don't know the story, and I have no knowledge that leads me to believe one way or the other.

I just naturally am a bit more skeptical of comments made about companies from employees who were fired. People at a company can certainly make untrue comments, and people who were fired can certainly make true ones.

It's just that, in my experience, there's a higher chance of negative bias in someone who was let go from a company.

True, should have mentioned that (edited it in). Relatively certain it was afterwards.

> if you completely do away with job titles, the politics would go away

For anyone who finds this plausible, it might be enlightening to contemplate (1) the first "hacker koan" at http://en.wikipedia.org/wiki/Hacker_koan and (2) this classic XKCD cartoonhttps://xkcd.com/592/ .

Plenty of people are still status driven. So either you have to decide you want 0 of these people at your company, or else ...

Honestly, my experience has been that structure is one of the key things that makes teams effective (not structure as in "rigid meetings/etc", but structure in the sense of "understanding what their role is and where they fit in")

This doesn't strictly require managers, of course, but self-organization rarely works out that well either as the org gets larger (you have the same problems, they are just hidden and nobody is accountable) :)

Sorry, but that's not how large groups of humans self-organize. Many idealists have tried, but power vacuums don't stay for long, nor are they efficient.

Yes, I am speaking from experience directly with a holocratic organization.

Thanks for replying, I was genuinely curious.

This reminds me of Jo Freeman's excellent essay, "THE TYRANNY of STRUCTURELESSNESS". Based on her experiences within unstructured feminist organizations, I think her insights are broadly generalizable.


Thank you for posting this, I hadn't read it until just now. It's clearly written, and the subject matter is historically accessible; I'm surprised I haven't seen it referenced in other discussions about Holacracy, Valve, etc.

Now that is a one-sided article if I've ever read one. Not to say that her position wasn't frustrating- I imagine that it was incredibly frustrating- but yeah OK she gets fired from Valve and then immediately goes on to do exactly what she was doing at Valve with the same engineers at a spin-off company that has Gabe's blessing producing stuff for Valve. How is that a loss?

It sounds way more like Valve wanted to do something that was a radical departure from anything they'd ever done before (hardware), the intrinsic structure of Valve meant that it was difficult if not impossible to set hardware up to succeed (peer-evaluated hiring, bonus tied to value of work), and then Valve realized it wasn't working and decided to make a clean break instead of drag it out. So then Valve immediately says "OK guys we couldn't make this work within Valve but we will back you up 100% if you want to go make this work external to Valve."

It's implied/addressed in the employee handbook. Basically the studio doesn't have fixed roles but roles are allowed to naturally emerge around projects and may shift at any time.

I don't think they are against people assuming responsibility or managing work, they just want to be more dynamic about it.

The valve employee manual is a great read. And well illustrated; worth a look: http://www.valvesoftware.com/company/Valve_Handbook_LowRes.p...

(From an earlier discussion at https://news.ycombinator.com/item?id=9250527)

> flat structure they have at Valve

is a lie. It all depends on who you know and are friends with. Read on how Valve handled Jeri Ellsworth and her hardware team, they were all fired with no notice by Gabe himself.

My experience working for one of "the big 3" automakers is that the very best employees take the buyout offer, confident that they can find a better job for better pay elsewhere. Years after the employee buyout programs began in the auto industry, they were requesting buyout packages of their own to save the company as a whole.

I believe this is the beginning of the "death spiral" described by PG in his "What happened to YAHOO" essay.

Not that bad. Best employees not always are the best "flat structure" employees.

And a lot of the great ones may want to hang on just for the fun of it.

Usually only the lower half of the stuff gets a "buyout offer" (strange name, BTW).

Every time this comes up, I have one simple question that I have yet to have answered to any degree of satisfaction.

(Someone who endorses this structure) explain to me in one sentence with fewer than 20 words how a 'lead link' is not the same role as a 'manager'.

I actually don't like the way they've defined lead link in Holacracy. It does feel too manager-like because they have the ability to create roles or remove people from roles.

We had this as a stepping stone at GrantTree, in our previous implementation of holarchy (note: that's not the same as holacracy, it's just one subset of holacracy that deals with how to organise work without organising people) but since then, we got rid of that managerial aspect (and also of the word "lead", which is needlessly managerial).

The links at GrantTree now do just one thing: they represent the circle they're a link of in the meetings of the supercircle.

This is just one role among many for the person, and it has absolutely zero managerial authority. It's an administrative and communications role. We're still in the early days of it, but the intention is that the person in that role will be rotated out every 3 months or so, to make sure the role doesn't turn into a manager role inadvertently (but at the moment there's no sign that it will).


A lead link can appoint people to roles, but the people have autonomy in that role. A manager has the authority to both appoint roles and micromanage execution.

Also, there is a parallel to lead link called rep link who is appointed by the lead link's circle to check against bad decisions/behavior by the lead link.

Another way to look at 'micromanage execution' is as 'support and mentoring in the role' based on honest feedback on perceived performance. How do people get support for challenges?

Providing support and feedback is a role responsibility. There can be as many responsibilities and roles to fulfill them as the team needs. If a circle member has a need for support and they are not getting it, it gets brought up at a tension meeting and quickly resolved (new responsibility assigned, role assigned, person's role changed, whatever's needed.)

Edit: I forgot to add that anyone who perceives a person is not receiving support for their responsibility can raise a tension. It's not just the manager's and the employee's job.

"If a circle member has a need for support and they are not getting it..."

How does circle member recognise that they need support? What indicators/who provides feedback?

My (limited) experience as a manager was mainly about nurture/support/challenge of colleagues. We could not allow them to fail/fail to 'ask' for support because of implications for clients (we did not sell people shoes).

The tension meeting makes it hard to avoid asking for support because it's so easy for people to bring up small problems as they're noticed and get quick resolution. Also, responsibilities are worded as objectively and measurably as possible (like good KPIs) to increase transparency into how people are doing in their roles. If a circle in an org like yours needs to avoid showing weaknesses to clients and to suss out problems before they're noticeable, they would be able to define those as responsibilities and assign them to the right people.

That said, I think one of the bets holocracy makes is that a group of people with a better understanding of their responsibilities and who receive quicker, clearer feedback will naturally be more proactive about problems. A circle is pretty unlikely to assign themselves the kinds of restrictive processing and reporting that are traditionally used to control interaction with clients in organizations that don't have perfectly reliable people. But in the short term, the control method would probably have more consistent results and that is not acceptable to every organization.

"I think one of the bets holocracy makes is that a group of people with a better understanding of their responsibilities and who receive quicker, clearer feedback will naturally be more proactive about problems."

Who provides the feedback? Based on what credentials? Explain why that person not a manager.

Tensions, which are problems or issues or defects, are provided by anyone who notices.

In addition to that, some roles may have responsibilities to provide particular support and feedback. For example, in a team of developers, a developer who has a code reviewer role may have the responsibility of providing feedback on others' code.

Other kinds of feedback-providing responsibilities could be proofreading, listening to sales calls, QA on manufactured items, feedback on negotiations, etc. Anyone could be assigned these roles/responsibilities.

Traditional managers usually see these responsibilities as their job and other employees don't easily get involved in the manager-employee feedback relationship.

"Tensions, which are problems or issues or defects, are provided by anyone who notices."

Interesting but 'deficit model' approach. No mentoring available for performance that does not cause what you define to be a 'tension' but that could be improved with little effort. I'm thinking of double-loop learning (Argyris) as opposed to single-loop learning which I interpret as covering proof reading/QA/Sales call monitoring.

Yes, I haven't seen anything requiring scrutinizing assumptions as a core role/responsibility, so it would have to be something the organization wants to do. I do think the model empowers employees to question assumptions very nicely once it is their responsibility, though.

If they also have the authority to remove people from roles, how do they not also have the authority to micromanage execution, at least not indirectly?

They would have to go out of their way to do it. It would be brought up as a tension and if people were unhappy with the link lead, the rep lead would bring it up to the link lead's higher-up circle (rep lead is a circle-elected peer of the link lead.)

So, instead of having direct authority to micromanage, they only have indirect possibility of micromanaging and there are explicit checks built in to prevent it.

>in one sentence with fewer than 20 words

Seems like an arbitrary restriction to put on something that's probably complex.

It's essentially "Explain it like I'm 5", which in this case can help discern between an interesting corporate experiment and PR bullshit.

No one wants to get halfway through an article claiming the former and ending up being the latter. If Zappos has a unique take on how to replace middle management, let's all hear it.

The ability or inability to describe an idea in simple terms has no relation to the validity of the idea.

>If Zappos has a unique take on how to replace middle management, let's all hear it.

I think their idea is in this book, mentioned in the article, that they asked all zappos employees to read: http://www.reinventingorganizations.com/

>The ability or inability to describe an idea in simple terms has no relation to the validity of the idea.

Maybe not in physics, but in politics and business it does - those fields like to seem more complex than they are so they fake it, which is much more difficult to do with simple language.

"Just say no to drugs" - simple terms. Is it a valid social policy given other ways to reduce drug use?

"From each according to his ability, to each according to his need". Simple terms. Does it lead to a workable political system?

"Three-strikes law". Simple idea. Does it lead to good social policy?

"Zero tolerance." Another simple idea. Does it lead to good social policy?

"White man's burden." Does it justify the American colonization of the Philippines and other countries?

"Curse and mark of Cain" as used by many religious groups to explain the lower social standing of those with black skin. Valid idea?

"Free silver at a ratio of 16 to 1" - one of the famous slogans of the Populist movement in the US in the late 1800s. Easy to say. Good economic policy?

Or do you prefer "Bad money drives out good" - Gresham's Law - as your economic policy?

Gersham's law is actually fairly valid, as well as being simple. The difference is that it isn't a statement of intent, it's a description of an inevitability.

I wanted to include more examples of short and fairly valid sayings for politics and economics. It was the only one I came up with.

A therefore B does not mean not A therefore not B

My point was that "A therefore B" isn't useful if the noise level is so high that it can't be used to make a meaningful decision.

Otherwise the advice reduces to "don't trust long explanations". But if 1% of long explanations are valid and 2% of short ones are valid, then the statement is both true, and worthless as a way to detect validity.

I disagree. I think we have more understanding of the physical world (physics) then we do the complex human interactions that make up society (politics) and the economy (business). To me the fact the latter fields are not well described alludes the them being more complex.

Everything can be initially described by an elevator pitch. That's never the whole explanation, it's just a sense that there's something in the idea (or not), and allows the listener to decide whether it's worth pursuing the long discussion.

It's not a guaranteed indicator, but the chances are better that a good elevator pitch is good because it describes something good.

If it can't be properly explained in simple terms, it might not be a great idea.

After you've done that, please explain in one sentence with fewer than 20 words how a determinant is a measure of fullness-of-span in n-dimensional space. I know that it is, but I've never understood the intuition.

V. I. Arnold uses 17 words:

"The determinant of a matrix is an (oriented) volume of the parallelepiped whose edges are its columns."


A determinant describes changes in volume under transformation, and an (n-1)-dimensional set of points has zero n-dimensional volume.

(very sketchy, but it's the best I could come up with under the constraints)

Brilliant effort!

I just finished Tony's book "Delivering Happiness". I admire how positive and optimistic he is. And I even agree with forcing self improvement on the employees.

But he also stresses running the company like a family aka cult. One of the advantages of moving to Vegas from SF was the employees were forced to spend more time together. He even said Zappos is his tribe. I wonder how these experiments would fare at a larger company.

Yeah, I read this article and for whatever reason had a sense of this being very cult-like. The terminology being used is similar to how cults load the language to make everyone feel like they are part the in-crowd. And, his offer to pay severance to anyone who wanted to quit just felt like how a cult leader will say "the door is always open if you want to leave."

Of course, he isn't actually building a cult, but I find the similarities of indoctrination very interesting.

They are provenly-effective techniques to make a disparate group of people work towards a common goal...

Whether the techniques and the end-goal are good or bad is left as an exercise to history :)

Strange it isn't former CEO as you wouldn't have a CEO in a flat company. Seems hypocritical to me, but maybe it isn't and I don't get the distinction.

" "All animals are equal, but some animals are more equal than others"

Just because you don't need management doesn't mean you don't need leadership.

Right, "nobody's job is essential...well except mine, of course."

Good managers lead.

Not every leader is a good manager.

Companies are required to have a CEO, as (famously) Gabe Newell does.

Zappos doesn't; it's a wholly-owned subsidiary of Amazon. They could all just be subordinate to Jeff Bezos.

Does the fact that it's a subsidiary mean it's not a company?

Otherwise, what separates a subsidiary from just being another team? I assume the difference is that a subsidiary can be legally considered to be a company on its own altogether, but I'm not a lawyer. If it is though, then assuming companies need a CEO, Zappos still needs one.

Depends. Someone has to sign the contracts. And to be the top dog legal wise. We don't have good legal framework that endorses flat corporations. Yet.

That's still rather different than a company with a hierarchy several levels deep.

I am wondering this too. Is it flat with a handful of executives on top?

Someone has to talk to the investors though, right?

bill in maintenance will take on that role.

Offering money for employees to leave is an interesting strategy. I worry about unintended consequences, though. One consequence is that the very best people (who have easy times finding jobs or striking out on their own) will be much more likely to leave than less competent people who may feel they would have a harder time on the open job market.

This sentiment is echoed a lot in relation to Zappos' policies, and I think it holds water. But it crossed my mind that the existence Dunning-Kruger effect might actually result in the inverse.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact