How Amazon Pressures Out 6% of Office Workers (seattletimes.com) 267
An anonymous reader quotes a report from Seattle Times, written by Katherine Anne Long: Amazon systematically attempts to channel 6% of its office employees out of the company each year, using processes embedded in proprietary software to help meet a target for turnover among low-ranked office workers, a metric Amazon calls "unregretted attrition," according to internal company documents seen by The Seattle Times. The documents underscore the extent to which Amazon's processes closely resemble the controversial management practice of stack ranking -- in which employees are graded by comparison with each other rather than against a job description or performance goals -- despite Amazon's insistence that it does not engage in stack ranking. The documents also highlight how much of Amazon's human resources processes are reliant on apps and algorithms, even among the company's office workforce. And they provide the most detailed picture yet of how Amazon uses performance improvement plans to funnel low-ranked employees out of the company. The company expects more than one-third of employees on performance improvement plans to fail, documents show. Amazon has previously said that its performance improvement plans aren't meant to punish employees.
The policies described in the documents reviewed by The Seattle Times apply to the company's office workforce, who comprise a minority of Amazon's roughly 950,000 U.S. employees. Amazon's warehouses replace workers much more frequently, The New York Times has reported: Before the pandemic, annual turnover rates at Amazon warehouses reached 150%. Amazon said some of the documentation reviewed by The Seattle Times was not created by the company's central human resources team and contains outdated terminology. But it did not dispute that the documents describe Amazon's internal policies. An Amazon spokesperson also said characterizing its performance management system as stack ranking is inaccurate. "We do not, nor have we ever, stack ranked our employees. This is not a practice that Amazon uses," said spokesperson Jaci Anderson, in an email. She said the goal of the company's performance review process is to "give employees more information and insights to continue to grow in their careers at Amazon."
Experts familiar with Amazon's processes disagreed with the company's stance that it does not stack-rank employees. Previous reporting by Business Insider has also found that Amazon grades employees on a curve. Amazon's performance-review system "forces [the company] to find the flaws in people as opposed to looking at their strengths," said longtime tech industry recruiter Chris Bloomquist, co-founder of Seattle's The Talent Mine. "If I have 10 brilliant people, but the least-brilliant person is fireable? That's stupid." The company's insistence that it does not practice stack ranking is "a bold-faced lie," Bloomquist said.
The policies described in the documents reviewed by The Seattle Times apply to the company's office workforce, who comprise a minority of Amazon's roughly 950,000 U.S. employees. Amazon's warehouses replace workers much more frequently, The New York Times has reported: Before the pandemic, annual turnover rates at Amazon warehouses reached 150%. Amazon said some of the documentation reviewed by The Seattle Times was not created by the company's central human resources team and contains outdated terminology. But it did not dispute that the documents describe Amazon's internal policies. An Amazon spokesperson also said characterizing its performance management system as stack ranking is inaccurate. "We do not, nor have we ever, stack ranked our employees. This is not a practice that Amazon uses," said spokesperson Jaci Anderson, in an email. She said the goal of the company's performance review process is to "give employees more information and insights to continue to grow in their careers at Amazon."
Experts familiar with Amazon's processes disagreed with the company's stance that it does not stack-rank employees. Previous reporting by Business Insider has also found that Amazon grades employees on a curve. Amazon's performance-review system "forces [the company] to find the flaws in people as opposed to looking at their strengths," said longtime tech industry recruiter Chris Bloomquist, co-founder of Seattle's The Talent Mine. "If I have 10 brilliant people, but the least-brilliant person is fireable? That's stupid." The company's insistence that it does not practice stack ranking is "a bold-faced lie," Bloomquist said.
PUP = Paper Trial For Firing Someone (Score:5, Insightful)
" And they provide the most detailed picture yet of how Amazon uses performance improvement plans to funnel low-ranked employees out of the company. "
Using "performance improvement plans" as a preparation for firing someone is pretty much standard across corporate America. The only purpose of these plans is to create a paper trail that can be used to show that a firing was for performance reasons and not for some legally prohibited reason (i.e. discrimination, retaliation from sexual harassments claims, whistleblowing etc.). Anybody with two brain cells should know that if you get put on a such a plan, it's time to look for a new job immediately.
Re: (Score:3)
I would agree, but I was stack-ranked out of a job in 2014 just because I was the least senior of my team. That was the *ONLY* reason.
To say nothing about the fact that I successfully worked for that same company for 7+ years in the late 1990s and early 2000s. I was re-recruited in 2013, but my seniority and net worth to the company was ignored by their stack-ranking process.
At least I got a severance that time.
Stack-ranking is stupid. General Electric proved that fact but it took too many years for every
Re:PUP = Paper Trial For Firing Someone (Score:5, Interesting)
Stacked ranking was pioneered by General Electric’s CEO Jack Welch, one of the worst CEO scumbags ever to stink up the place. That in itself should be a warning.
Stacked ranking stifles innovation and fosters artificial competition between employees, Everybody becomes so preoccupied with their ranking that they lose sight of the real goals and waste time on snitching on their co-workers and covering their own asses.
I have seen it in use at past employers like Intel and HP, both of which had thoroughly mediocre IT staffs.
It's a race to the bottom.
Re:PUP = Paper Trial For Firing Someone (Score:5, Insightful)
PIPs are for employees performing poorly and not meeting expectations. Most employees performing poorly continue to perform poorly despite the PIP. Then they lose their job. But it isn't because of a vast conspiracy. It is because they didn't improve.
Only if you assume that the PIP is evaluated objectively by someone other than the manager who put the person under a PIP. Otherwise, it's basically just a way to check off checkboxes on a legal compliance worksheet, and can be used as cover for terminating somebody for any arbitrary non-performance reason, up to and including outright discrimination.
Re:PUP = Paper Trial For Firing Someone (Score:4)
PIPs are for employees performing poorly and not meeting expectations. Most employees performing poorly continue to perform poorly despite the PIP. Then they lose their job. But it isn't because of a vast conspiracy. It is because they didn't improve.
Only if you assume that the PIP is evaluated objectively by someone other than the manager who put the person under a PIP. Otherwise, it's basically just a way to check off checkboxes on a legal compliance worksheet, and can be used as cover for terminating somebody for any arbitrary non-performance reason, up to and including outright discrimination.
I have seen them used both ways. Mostly as a tool to get rid of someone that a manager took a disliking to. When you work near enough those people you can see the disconnect between what the manager is saying and what the employee is doing to see how it is being abused. But, I have also seen it used to funnel people who were utterly incompetent out of the company. The most recent example was a person who used to have an army of contractors that he effectively managed. Turns out when the army of contractors were removed he had no actual skill. Apparently he just rebranded the contractors work as his own.
Re: PUP = Paper Trial For Firing Someone (Score:2)
Re: PUP = Paper Trial For Firing Someone (Score:5, Funny)
Office Space was a documentary.
Re:PUP = Paper Trial For Firing Someone (Score:5, Interesting)
The most frustrating thing about reviews was how many managers would sink your entire year around one or two isolated incidents, regardless of how much else you achieved. I had so many colleagues, myself included, who would get dinged for having a disagreement with a co-worker, or just upsetting someone and they mention it to your manager.
Re:PUP = Paper Trial For Firing Someone (Score:5, Insightful)
The most frustrating thing about reviews was how many managers would sink your entire year around one or two isolated incidents, regardless of how much else you achieved. I had so many colleagues, myself included, who would get dinged for having a disagreement with a co-worker, or just upsetting someone and they mention it to your manager.
Yup. And that's because when you're in a stack-ranked or almost-stack-ranked (performance quotas) organization, somebody has to be the fall guy even if everybody had an amazing year. At some point, you just say, "screw this" and leave for another team where the management are willing to stand up for their employees.
Re: (Score:3)
Notice how those rank and yank cube farm hellholes have slid down the desirability rankings. Google is now way below Microsoft and Microsoft is hardly a plum assignment.
Re:PUP = Paper Trial For Firing Someone (Score:4, Informative)
Notice how those rank and yank cube farm hellholes have slid down the desirability rankings. Google is now way below Microsoft and Microsoft is hardly a plum assignment.
In Denmark Microsoft does in fact have a reputation as an excellent workplace. They (and subsidiaries) have been rated the best (50+ employee) IT workplace in Denmark a number of times (not sure how many, but they did place first three years in a row at one time). But it may of course be different in other countries.
Re:PUP = Paper Trial For Firing Someone (Score:4, Funny)
Just give it time. Getting to know Microsoft is getting to hate it.
They have had offices in Denmark for over 30 years.
Re: (Score:3)
I have seen them used both ways. Mostly as a tool to get rid of someone that a manager took a disliking to. When you work near enough those people you can see the disconnect between what the manager is saying and what the employee is doing to see how it is being abused. But, I have also seen it used to funnel people who were utterly incompetent out of the company. The most recent example was a person who used to have an army of contractors that he effectively managed. Turns out when the army of contractors were removed he had no actual skill. Apparently he just rebranded the contractors work as his own.
Umm... I would argue that managing an army of contractors *is* actual skill, just not the right skill set for the position as redefined without the army of contractors. So that's more like a layoff than terminating an incompetent employee. :-)
Re: (Score:3)
Where I work, management doesn't want a huge headcount since it looks bad on them (somehow). So we have always had an extremely high contractor to fixed hire ratio (i.e. 3 internal devs to 20 external). This means the fixed hire developers' time is spent on defining the work, talking to business and explaining to the contractors what to do. Most/all the fixed hires are now very rusty developers and likely would have trouble developing very
Re: (Score:2)
PIPs are for employees performing poorly and not meeting expectations. Most employees performing poorly continue to perform poorly despite the PIP. Then they lose their job. But it isn't because of a vast conspiracy. It is because they didn't improve.
Only if you assume that the PIP is evaluated objectively by someone other than the manager who put the person under a PIP. Otherwise, it's basically just a way to check off checkboxes on a legal compliance worksheet, and can be used as cover for terminating somebody for any arbitrary non-performance reason, up to and including outright discrimination.
Ultimately, that's the big problem.
There's nothing wrong with a company evaluating employees and getting rid of the ones who are not doing their job. That's basic common sense.
Unfortunately, the managers in charge of doing these evaluations have a tendency to turn into dictators and it becomes a very abusive process.
Re:PUP = Paper Trial For Firing Someone (Score:5, Insightful)
PIPs are for employees performing poorly and not meeting expectations. Most employees performing poorly continue to perform poorly despite the PIP. Then they lose their job. But it isn't because of a vast conspiracy. It is because they didn't improve.
Only if you assume that the PIP is evaluated objectively by someone other than the manager who put the person under a PIP. Otherwise, it's basically just a way to check off checkboxes on a legal compliance worksheet, and can be used as cover for terminating somebody for any arbitrary non-performance reason, up to and including outright discrimination.
Ultimately, that's the big problem. There's nothing wrong with a company evaluating employees and getting rid of the ones who are not doing their job. That's basic common sense. Unfortunately, the managers in charge of doing these evaluations have a tendency to turn into dictators and it becomes a very abusive process.
There's a cure for that in several parts:
Do that, and those problems mostly go away.
Re:PUP = Paper Trial For Firing Someone (Score:4)
I was put on a PIP once that was an obvious front for firing me because one of the bosses didn't like how I called out his complete lack of leadership. I was nice about it, but we had no direction, the project was treading water, and one of my bosses said that he was just 'making up tasks to do' as if that was a good thing. Not only did management later send out an email apologizing to the whole team for the lack of direction, the project was eventually cancelled. They fired me before it was cancelled, though. It's fine. I took the severance and found a better job.
At the new place, I was put on a PIP because I was legitimately put into the wrong place. We were trying to get a very struggling game out the door, and they fired the biggest people-hose I've ever seen. I got put onto a team that required fairly specific low-level console network knowledge, but nobody had the time to train me or show me how to do anything. I treaded water for a while before telling my boss that I didn't think I was in the right place. He agreed, chastised me for not bringing it up sooner, and put me on a PIP because my performance had been legitimately underwhelming. But once I was on a team where I actually knew what I was doing (and the people-hose had been withdrawn and the ship date delayed, because throwing more people at struggling projects absolutely never works well), I had normal-to-excellent reviews again.
Sometimes it's not the employee that's the problem, it's the situation that the employee is put in, and that's a failure of management. I've never been put on a PIP again, or performed below expectations. There are some legitimately bad programmers out there, but my own experience is that good programmers put onto the wrong teams or forced to work on systems that they don't want to will ultimately turn in poor results, and it's management's fault for wasting that time, money and talent. I've met far more bad managers than bad programmers.
Re: (Score:3)
Been there, done that. Excuse for a firing. Been on a PIP twice, same company. Both times, the first indication that there was a 'problem with my performance' was being called into the manager's office for the PIP. Both times, several pertinent questions I asked were left unanswered, several extenuating circumstances ignored. Both times, the "goals of the PIP" were well in advance of any previous expressed expectations.
The first time, due to clever (and fortunate) choosing of tasks, and some 12/7 weeks
Re: (Score:3)
one of the bosses didn't like how I called out his complete lack of leadership
"calling out" your bosses is, by definition, insubordination. If you want to punch up, you can be expected to eventually be fired. You're lucky they made an attempt to reform you first, and gave you time to quit.
If you think the boss exhibits a "complete lack of leadership," why are you working there? Do you want the job, or do you just think this company is required to give you a paycheck?
Re: PUP = Paper Trial For Firing Someone (Score:3)
Depends how you do it. If it's mutiny-style with the whole team watching and cheering, then yes, it's subordination.
If it's a constructive feedback, preferably in private, it's not. Your superior can still choose to take offence, but then he/she's a dick.
As for the "why are you still workimg there"-part... well, constructive fredback goes both ways. Your boss is just a person, not some magical alien race. They go home at the end of the day, broken or happy, tell their spouse about how the day was, and try j
Re: PUP = Paper Trial For Firing Someone (Score:3)
That was figuratively speaking.
But even if it wasn't, you're wrong. There's dynamics of the mob involved. Everyone has a beef, however small, with the manager. If the issue is even a bit controversial, it's essentially impossible to discuss with more than 3 people involved, even more so when there's a power gradient involved. Discussion may deviate into the unconstructive, and then it's often impossible - even afterwards - to fix it. Many times this inevitably leads to someone having to go. Knowng or fearin
Re: (Score:3)
The literal definition ... (Score:5, Insightful)
... of a rat-race.
Corporations are psychotic by their very nature.
Do not trust corporations. Not as customer and not as employee.
Re: (Score:2)
Did anyone complain about a rat race when AMD delivered Zen 3 CPUs that performed better than what Intel had? I've also really loved the scramble by the electronics manufacturers to get me to buy a 4K TV as well. Maybe we'll start getting a race to deliver a cheaper electric car. Sure Tesla has mov
Re: (Score:2)
Re: (Score:3)
Sure that's all well and good, but at the end of the day most of us are going to do business with whoever delivers fastest and at the lowest cost. If we didn't there'd be no incentive to deliver better service.
But does this practice actually produce the best results overall, or does it create a constant churn of lower skill employees as the original bottom 6% might have been acceptable doing their jobs, but have long since been replaced by people who will possibly also be replaced? Does this stress of being in the bottom 20% force other adequate workers to find employment elsewhere, causing those lower on the rungs to move up, in a perverse peter-principle parallel? To say nothing of training a new crop of 6% o
Re:The literal definition ... (Score:5, Informative)
... of a rat-race.
Corporations are psychotic by their very nature.
Do not trust corporations. Not as customer and not as employee.
No, it's not the literal definition of a rat race, it's the figurative definition of a rat race. The literal definition of a rat race involves at least 2 actual rats competing to reach a goal first.
Re:The literal definition ... (Score:5, Insightful)
No kidding.
HR dept exists to insulate and protect the company from employees. Avoid going to HR to complain about something at all costs - chance they'll turn it around on you, its a calculated risk.
Nod, smile, stay out of politics, avoid having an opinion, dont react to the jerk employees or get sucked into what ever their "issue" is, and just get your work done.
Avoid doing wok that there wont be a record for, e.g. people not logging tickets or going through the right intake channels. Otherwise you're hiding work and lowering your measured output should anyone start pulling reports out of those systems to see who's doing what.
Dont trust those internal employee feedback surveys, just click favorable ratings. They say they're anonymous but I don't buy that.
There are some exceptions, I have worked for managers that are actually human and care, but be wary.
The OG Stacked Ranking (Score:3)
"We're adding a little something to this month's sales contest. As you all know, first prize is a Cadillac El Dorado. Second prize is a set of steak knives. Third prize is you're fired."
In and Out, two sides of the same coin. (Score:5, Interesting)
Re: (Score:3)
It's called raising the bar. You need to raise the bar (be better than) 50% of the people doing the same job in key areas of the position.
As for the stack ranking, it's simple. You take everyone and rank them, then split it up into the top 13%, bottom 7% and everyone else in the middle. if you are in the bottom 7% you get a PIP. The manager of this person is encouraged to either help the person who is on the PIP to move on or not be in the bottom 7% next time otherwise, the manager is automatically put on a
Re: (Score:2)
he was very clear that they had a policy to ensure every hire was above the 50th percentile compared to all employees
That's a nice quote but unlikely to be reality.
Indeed. Over the long run, the policy fails exactly half the time.
Re: In and Out, two sides of the same coin. (Score:2)
Re:In and Out, two sides of the same coin. (Score:5, Informative)
You obviously don't know anyone who works at Amazon then. I've heard this same thing from at least 12 different people that I know who work for Amazon/AWS. And Netflix is bad for this as well. It's their policy and a reality.
Comment removed (Score:5, Insightful)
Re: (Score:2)
I usually find out after I've already signed on and left my old job. Watching your company hand out 1 PIP per team every other quarter is rather disruptive and demotivating.
Re:Interview (Score:4, Informative)
No sane company keeps low performance people, true. But, there are ways to manage performance that aren't stack ranking. Stack ranking says that if I have 10 equal performers on my team, and all are high performers, I have to find a reason to call one of them a low performer and likely to put that one on a PIP or outright get rid of that person. When I saw this in practice, the better managers would sit their team down and say "look, this sucks, and it's not right, but someone has to take the hit. So we're going to rotate it through the team, and everyone will take the hit one quarter." The best managers would stand up for their team to upper management, and quickly found themselves in non-management roles because of it.
Stack ranking breeds a toxic culture of back-stabbing and mistrust where people look for ways to point out the mistakes of others, instead of supporting the team and helping each other grow and succeed.
I'm very happily a former employee of that company.
Re: (Score:3)
Yep, I've had to do this far too often. Picking the worst person on a high-performing team is painful and hurts the team overall. I had 5 great performers who exceeded all goals placed in front of them. When I explained this to the director, I was told that it's my fault they all succeeded because I didn't set the bar high enough. When I pointed out that the goals I set were simply the goals he gave my team...well let's say my review said I wasn't a "team player" and my bonus suffered. In the end, I picked
Forced quotas are stupid (Score:5, Insightful)
I suspect that Bezos knows this on some level, but simply doesn't care.
Re:Forced quotas are stupid (Score:5, Insightful)
It tells you a lot about Amazon's attitude towards employees. They are not human beings who respond when you help them to get the best from themselves. They are robots and you cull the weakest 6% every year, just like you throw out your old laptops.
Re:Forced quotas are stupid (Score:5, Insightful)
Any kind of forced quota system is stupid. Maybe you think you're getting rid of the worst 6% of your workforce, but that's really relying on your metrics being good (they aren't good enough if they're good at all) and that none of the employees will game them (good luck with that) or subvert the system. Even if that were mostly true, you're also losing all of the good workers that don't want to deal with that kind of bullshit along with all of the people who will never apply to work there because of that bullshit.
Anybody who thinks stack ranking is a good idea is a fool, period. The reality is that people don't usually just underperform because they're bad employees. Most underperformance is the result of bad management, statistically speaking. And when it isn't, it's usually the result of something going on in the employee's life that the management doesn't know about. In fact, it is often for reasons that would make the person part of a protected class if that person reported the details to the management.
In other words, stack ranking is basically a way to sneak around labor laws and punish workers for things that are mostly beyond their control. Life's too short to work for a company that thinks this is a good idea, because you never know when you're going to have a bad year for whatever reason and end up in that bottom 6% or whatever. It happens even to the best of us. And at some point, it also starts to look an awful lot like age discrimination.
And I can't imagine why any CEO would think that it's a good idea, either. Stack ranking actively encourages employees to leave the company for better offers. After all, why would anyone be loyal to a company who they know won't be loyal to them if they need it to be?
Re: (Score:2)
Managers are also reviewed based on the performance of their team. If one manager has a bunch of low performing employees, it calls in to question both their ability to hire and their ability to lead.
Stack ranking is not necessarily bad. You have to have some means to determine relative performance to determine raises, promotions, bonuses, etc. Not just for clearing the chaff. The one area it fails is when it requires the bottom performer on each team to be replaced. That hurts when there are small tea
Re:Forced quotas are stupid (Score:5, Insightful)
Managers are also reviewed based on the performance of their team. If one manager has a bunch of low performing employees, it calls in to question both their ability to hire and their ability to lead.
Stack ranking is not necessarily bad. You have to have some means to determine relative performance to determine raises, promotions, bonuses, etc. Not just for clearing the chaff. The one area it fails is when it requires the bottom performer on each team to be replaced. That hurts when there are small teams that are all top performers, but a large team has several poor performers. Team by team performance has to be weighted.
It fails in a lot of ways, actually. Replacing people should be based on a pattern of actual poor performance over a long period of time, and should be preceded by getting that person to try a different team that might be a better match for the employee's skillset. Otherwise, you're just getting rid of good people for accidentally ending up in a bad team.
And stack ranking is inherently unfair for determining raises, too, because you're ranked relative to your team, not relative to the whole company (unless your job is easy to evaluate objectively), which means you can be an amazing performer, and if everybody on your team is also good, you end up with fewer raises than if you switched to a team with fewer smart/competent people. It leads to the mediocritization of the entire company, without any "tiger team" groups that can solve problems quickly, because nobody would want to be on that team, knowing that they would never get a raise again.
I literally can't think of any advantages of stack ranking over simply evaluating each employee based on that person's achievements without trying to compare it to another person. Creating software is not a competition. If you're competing against your coworkers, this creates an unhealthy work environment, and eventually leads to idiotic silo problems as well, where one team doesn't want another team to do well because it will make them look worse by comparison. And then suddenly you're creating your fiftieth chat app or whatever because each team wants to beat the other team.
Stack ranking puts the focus on competition, rather than collaboration. And that is inherently harmful in and of itself, even if it didn't cause other serious fairness problems.
Re:Forced quotas are stupid (Score:4, Insightful)
The harsh reality is when you set an arbitrary quota for "deadwood", you are going to be firing people who are not actually deadwood.
If you've got a team of 10 good people, but you gotta fire 10%, you're going to be firing a good person. This utterly destroys your work environment. The thing that starts to matter is your social skills and ability to manipulate, not your actual work output. The successful people are the ones best able to sabotage others or take credit for others work.
The people who are actually good? They quit in droves to get away from all the backstabbing.
150% annual turnover of ~1M employees (Score:2)
Hire to Fire (Score:5, Interesting)
Before you rush to post new comments, note that in May there was a submission about the 'hire to fire' practice by Amazon managers [slashdot.org]. Since they are forced to fire some people, they hire extra people with the only goal of firing them, instead of one of the good team members.
Comments in that link confirm and summarize all that is claimed here.
Re:Hire to Fire (Score:5, Funny)
Before you rush to post new comments, note that in May there was a submission about the 'hire to fire' practice by Amazon managers [slashdot.org]. Since they are forced to fire some people, they hire extra people with the only goal of firing them, instead of one of the good team members.
Comments in that link confirm and summarize all that is claimed here.
Of course, we rats quickly learn how to get the cheese rewards. Need to cut x% per year; hire x% extra so you can always meet the quota. I worked for a company that mandated 1 costs savings suggestion per employee per year. We'd go to the filling cabinet, pull one out from a few years back, copy it, white out the data and put this years, and send it in. My boss checked the box and we never heard back from HQ. Once I suggested firing the person who read those because they didn't o anything, the boss laughed but said no why ruin a good thing by replacing them with someone who might read them?
Re: (Score:2)
In other words, stack ranking is bad because it's leasing people to game the system.
Because honestly this is pathetic that people are having to work around the system lik
Re: (Score:3, Funny)
Explains why the new employees are given red shirts.
Good idea (Score:2)
Having worked in many companies with lots of "top" talent, culling 1 out of 20 (5%) every year or every other year is a great idea. There always is that guy/gal that 1) isn't actually that great (How is it that people from Berkley interview great, then...) or 2) is great, but isn't "focused" (read: lazy or mentally not all together) or 3) has a shit attitude - no matter how good they are.
Re: (Score:2)
It has a history that dates back to General Electric in the 1980s. https://en.wikipedia.org/wiki/... [wikipedia.org]
It has similarities with the "up or out" policy of business partnerships and the US military (https://en.wikipedia.org/wiki/Up_or_out#Military).
Speaking on for myself, I don't want to be kept in a job where I'm under-performing that badly. Jobs come and go.
Re: (Score:3)
The difference was what I was b
Re: (Score:3)
The problem with stack ranking is that, by definition, someone has to be deemed to be "underperforming" regardless of how well the team is performing overall. It's toxic and short-sighted.
Removing the bottom 10% (Score:2)
"If I have 10 brilliant people, but the least-brilliant person is fireable? That's stupid."
While I don't agree with the policy, this doesn't seem to be a valid analogy. It's not hard to find a group of 10 brilliant people. When you get to a group of 1000 (or even 100) it's unlikely that all of them are keepers. Even if you do have 100 brilliant people, what are the chances that they all work well together and fit with the company/team's vibe?
Be honest, sounds good (Score:2)
I don't think any company with more than 10 people in it, could not stood to have had 6% of it's people leave.
The sad truth is managers will not do what needs to be done, so maybe having process do it instead is a good idea.
Question is, will process get rid of the people who need to go or not?
Re: (Score:2)
I'd argue that the question is less "will it" and more "which process is best at it".
There are few things as demoralizing to a top 20% worker seeing the bottom 20% worker get the similar or same pay as you "because equity". That's how you get equity of performance with top tier performers rapidly to level close to bottom 20% in performance that Soviets observed in 1930s when they started implementing that exact policy and tried to fix with Stakhanovite movement.
The problem is that there's a lot of personal
The First Rule of Camping (Score:2)
Sounds like this applies pretty well to working at Amazon. You don't have to be the best, just better than somebody else.
were gonna need to move you downstairs to storage (Score:3)
OK, I really really need to know (Score:2)
If they manage to identify and "Pressure Out" more than 6% would HR/Management get more or less money?
I didn't know stack ranking was controversial (Score:2)
Most of the places I've worked at use stack ranking and get rid of the bottom 5%-10% annually.
The alternative is not paying them, which does them no favors.
Seriously, it isn't that hard to avoid being in the bottom 10%.
You have really screwed up when management perceives that 90% of your peers are better performers than you.
Re: (Score:3)
It punishes managers for building high-performing groups. And punishes employees in those groups. If you have to evaluate on a curve, you are bad at measuring performance, bad at hiring, or bad at goal setting. Probably all three.
Re: (Score:3)
You have really screwed up when management perceives that 90% of your peers are better performers than you.
Or you're one of 10 people on a fantastic team. Or you don't have the best social skills, so you've been out-maneuvered in the break room by people that do.
People who talk positively about stack ranking only think about it in the abstract and some magical land where it can be applied to the company as a whole. In reality, it gets implemented by every first-tier manager being forced to fire 10% of their team every year.
Re: (Score:3)
Good. By gaming the system I assume you mean by DOING THEIR JOB, right?
Not even close. Stack-ranking breeds back-biting chiseling little weasels who badmouth all of their team members regardless of how well they're doing, who steal credit for other people's work, and who make mountains out of molehills to make it seem like they've achieved great things. It is a perverse incentive structure which breeds bad behavior, and ultimately kills companies.
It motivates you not to fuck up, not do good work (Score:5, Interesting)
Good. By gaming the system I assume you mean by DOING THEIR JOB, right? Not everyone is equal and deserves a paycheck. You need to learn that.
Have you never held a job? Work for a big company and you will encounter politics. Stack ranking motivates you to not make mistakes, not do good work. There's a huge difference. Stack ranking motivates you to waste energy kissing ass of decision makers as well as shit on your peers, hoping they'll be the one fired and not you. As it is, I have seen a lot of assholes rise through the ranks by lying, blaming others, taking credit for work they don't do, etc. If you've ever worked a day in your life in a medium to large company, imagine the biggest asshole in power...now imagine how much bigger of an asshole they could be if they could get rid of people they don't like.
I have one of those in my company. He has a team of feeble engineers who don't threaten his ego. They don't do their job very well and cause lots of issues. He should be fired, but he blames someone else every time something goes wrong and apparently has figured out how to flatter his equally useless boss. Now imagine someone like that who has the power to fire anyone who reports to him and makes him look bad. Their error backlog is huge. They're the only team that doesn't write unit tests. The DBA is constantly having to call them because they don't know what an index is, how to design DB tables and are constantly causing table scans, slowing down the infrastructure for everyone. This loser is who the system is designed to prevent, but he's worked at stack rank companies before and gotten promoted...why?...he knows how to manipulate those in power.
By being better at office politics than engineering, people like this are able to thrive. They're given more power in stack ranking because now engineers are vulnerable and too scared to speak up. If you threaten to shit can me, I am going to focus on not getting shit canned, not what's best for the company or the work product...or, more realistically, go find a better employer. Stack ranking assumes it's not easy to find non-dysfunctional employers at the drop of a hat, like it is for business programmers these days.
A better alternative? Don't hire shitty employees in the first place...if they start great and turn out shitty, either put them on performance improvement plans or fire them...don't make everyone live in fear. This is not how top companies operate. If you want to do good work, you have to have some freedom to take risks and focus on medium and long-term success, not just short term success so you don't get fired.
Good idea, bad execution (Score:5, Insightful)
The basic idea is to each year fire the worst employees, never allowing a bad employee to gain any form of tenure. Also it tends to stop the Peter Principle, i.e. if you do your job well you keep getting promoted till you get stuck at the one thing you are bad at. It also lets the boss fire anyone they want to, as the firing methods are built in and hard to argue with.
The problems with it, is
1) They do not apply it to senior executives. This is compounded by the aforementioned Peter Principle.
2) The bosses know about this 6%, so they hire people for the sole purpose of firing them that year.
3) The stupider bosses will fire people for stupid or prejudiced reasons and get away with it. In fact, a stupid boss is MORE likely to fire a good employee, because they do not understand or like them.
4) Turnover is expensive. Training costs money, plus the first couple of months before you learn your job you are not great at it. Particularly if you were hired to be fired. Better to have longer term employees that are good at their job.
5) It tends to kill the concept of hiring from within. If you make it for a year or two, you know you know the job and your bosses like you, so your job is safe. Not so if you get promoted. Better to look for a new job at a new place than to try and get a promotion.
The Peter principle is stupid (Score:3)
On the other hand maybe this makes it easier for them to hire more h-1bs and less local talent.
If they were bad, why did you hire them? (Score:5, Insightful)
The basic idea is to each year fire the worst employees, never allowing a bad employee to gain any form of tenure.
Why are 6% of your employees worth firing in the first place? Stack ranking has been proven to be counterproductive. Sure, some percentage of employees who were great hires become HR cases that have to go, but if you're doing it right, it shouldn't be 6%. I'd guess 1% tops. It's been well documented that stack ranking systems create insecurity, bad work product, fear, stress, and internal politics.
Good work comes from creativity and confidence that if you take a risk, you won't get fired. This is a major reason why every famous billionaire came from a wealthy family, including Jeff Bezos, but also Bill Gates and Mark Zuckerberg. They could take wild risks and start companies because they knew that if they got in too much trouble, their family would bail them out. There are many smarter, harder-working, creative kids from poorer backgrounds, but you can't really start your own company if you're unsure if a mistake will make you homeless...so those brilliant people work for billionaires instead of becoming billionaires themselves. Similarly, employees, need to be able to take risks and try hard projects that could fail, like AWS or the Kindle.
Seeing my coworkers fired doesn't motivate me to do good work, it motivates me to not fuck up. There's a huge difference. In order to do great work, I need comfort. I need freedom. I need security. If you make me live in fear, I'm just trying to ensure I don't make anyone angry or miss something. I'm going to hedge my bets and play politics and you give me a motivation to shit on other people to ensure they're part of the bottom 6% and not me.
Stack ranking creates a hostile work environment. It's been well proven that it doesn't contribute to better work product. There are better ways to push and motivate your employees, hence why Amazon's superior competitors don't do this.
Stupid? (Score:2)
There are lots of reasons one might object to Amazon's system. You might feel it's cruel or creates an unpleasant work environment but one thing it isn't is stupid. Amazon is a huge company and it's not pulling low ranked office workers from some elite group so the idea that amazon might end up with only excellent workers (the 10 brilliant workers bit) is just ridiculous.
I mean I might not want to work at a company with that kind of policy and one might suggest that, all things considered, it's not even p
Funny (Score:2)
I'm surprised they need to force anything (Score:2)
I've known a handful of people who were employed by Amazon, although at somewhat higher job levels than what is described here. Given their descriptions of what it's like to work for the company, I would expect well over 6% of employees would happily choose to leave of their own volition during any given year.
Comment removed (Score:4, Insightful)
Too bad their quality doesn't reflect it. (Score:5, Interesting)
Generally, I tend not to opine on hiring practices. If you try to get rid of 6% of your workforce, it sounds like you suck at hiring. It also creates a hostile and political work environment..so hearing that, I am glad I turned down Amazon. However, the real evidence is your results. If Amazon was producing superior products to their competitors, I'd say "OK, that's not for me, but glad they found what works for them," but they're just a mess. Amazon is not an well admired company by it's customers. "Amazon quality" is not a compliment. They make lots of money..but so does McDonald's. It doesn't mean their workforce is talented and their hiring practices productive or well thought out.
Amazon just added a new dark pattern (Score:3)
https://www.amazon.com/Granola... [amazon.com]
Jack Welch and the "Vitality Curve" (Score:2)
You can thank Jack Welch [wikipedia.org] of General Electric, who invented the so-called vitality curve [wikipedia.org].
It may be justified during a reorganization, or when a company is under new management. But how can they justify doing it year after year for many years? See, if there is dead weight you cull it on the first year, may be the rest on the second year. Then what? It becomes just a toxic workplace.
Then you have companies like Amazon, where some managers hire people specifically to get around this policy, and have people on
amazon needs unions (Score:2)
amazon needs unions.
Unions may keep some poor workers but they also help stop BS and other stuff.
Re: (Score:3)
And on the flip side, you get what you pay for. A company which pays its employees no more than absolutely necessary may have all of the paperwork necessary to fire underperforming employees, but they can neither hire professionals nor improve the performance of their existing employees. If you treat your employees like imbeciles, they'll eventually take the hint. If you treat them like professionals, they'll rise to the occasion.
Underperforming employees are literally a management failure.
You eithe
Re: (Score:2)
If the bare minimum isn't good enough, then just raise the bare minimum so that it's acceptable. If they want you to wear 20 pieces of flair as the bare minimum, then they should just make 20 pieces of the flair the bare minimum.
Market does not guarantee your performance (Score:2)
Well, in an actually healthy free market, they would have an *interest* to keep you, since you could make just as many demands as them, since what you brought to the table is just as valuable.
How does an "actually healthy free market" guarantee you'll be any good?
Regardless of if a market is free or not, it's all on you to be an "actually good employee" who can do a good job.
That they can just take you, and throw you away like a drained sponge,
More like if you found a sponge that wasn't actually absorbing a
Re: So? (Score:2)
I actually watched that video, and at no point does it argue that it's fascist, let alone propaganda. Though that video has all kinds of glaring problems. First, the creator doesn't actually know what free market means, what he's actually describing is laissez-faire. Free market means that the prices are governed by supply and demand. That is, at opposed to some entity arbitrarily setting prices, a la socialism. Second, pretty much every point he makes is a false dichotomy, especially the whole coconut bit
Re: (Score:2)
Fascism involves private ownership of companies with government control.
What he said may be unrestrained capitalism with no safety nets for its rough edges, but it is not fascism.
If you seek a contemporary example of fascism, look to government threatening social media corporations with untold billions in stock losses via section 230 changes unless they censor harrassment -- hey! Our political opponents right before an election are posting harrassing tweets!
Re: (Score:3)
Keeping around poor performers also tends to result in your good performers leaving. When they see a bunch of people not performing, they tend to feel like they're the ones keeping the ship afloat. Top performers want to be around other top performers and feel like everyone is doing their fair share of work.
Re:So? (Score:5, Insightful)
Keeping around poor performers also tends to result in your good performers leaving
What do you think happens in year 2 or 3?
You already fired the poor performers. But you're going to fire another 10%, because you decided that's the way to get rid of all that "dead weight".
The result is you start firing people based on their social skills, instead of work product. Because you already eliminated the workers that are actually bad. Your workers figure out very quickly that the only way to be secure in their job is to go to war with their coworkers, and actively sabotage other people to lower their perceived rank.
Now those "top performers" quit because they'd rather not be in an environment where everyone is constantly stabbing each other in the back. And you're left with a company of mediocre sociopaths.
Re:So? (Score:5, Interesting)
There were another 100 where were also pretty bad, and wouldn't be missed. 90 of them will be gone the second year.
Ok, now do year 3. Or 4.
When you follow this strategy, you very quickly reach a point where you are firing people who are not bad in order to hit your pulled-out-of-the-air number. The remaining workers will not happily work away, counting on chance to keep their job. They're going to stop collaborating and they're going to start sabotaging in order to keep their job.
Your top performers aren't happy the "deadwood" is gone, they're pissed off by working in an incredibly toxic environment where politics is the most important skill. Or they're on a team full of top performers, and now have to worry about getting fired themselves to meet your metric. And since they're quite good, they quickly find a replacement employer who doesn't want to only employ sociopaths.
Re: (Score:3)
You're only thinking about year 1 or year 2.
What about year 3? You've already gotten rid of all those employees that "don't perform", but you gotta fire another 10% because you're stack ranking.
So now you're firing the roughly 1% crappy people who sliped through your interviews, and 9% decent people.
And then you get to year 4. By now, the remaining employees have figured out the only way to stay employed is to actively sabotage your coworkers. Because they're also "decent", and that's not going to protec
Re: (Score:3)
Wait, wait, what if in year one you fire 10 people and also hire 10 replacements.
Then in year 2 you fire 10 people and also hire 10 replacements.
Then in year 3 you fire to people and also hire 10 replacements.
There is no shrinking pool, there is not some smaller number of "remaining employees." There is the same number you started with. And good employees are getting better with experience, after 3 years they have much less to worry about. If they're got 3 years of experience and performing below new hires.
Re:So? (Score:5, Insightful)
Wait, wait, what if in year one you fire 10 people and also hire 10 replacements.
The assumption behind stack ranking is 10% of workers are terrible. So you fire 10, and hire 10 new ones. Now you only have 1 terrible worker, but will be firing 10.
And good employees are getting better with experience, after 3 years they have much less to worry about
Why? You only have 1 person who actually needs to be fired. You're going to be firing 9 good people to hit your firing target. So those years of experience aren't at all helpful since management is having to hunt for someone, anyone to fire. Plus greater responsibility means greater exposure when your coworker sabotages you to avoid being in the "lucky" 9 themselves.
you should fire them, because it is really bad for the morale of the high performers to keep those people around.
You know what's way, way, way worse for the morale of the high performers? Having to work in an environment where everyone is backstabbing everyone else to save their job. Runner up: watching another high performer being fired because it's a team of 10 and someone has to get fired, even when they're all good.
Re: (Score:3)
You're getting the premise wrong, because you're getting it by waving your hands at the problem instead of listening to the people who advocated the system, or the people who use it.
The premise is that a certain amount of turnover is beneficial. For a basket of complicated reasons. You're waving your hands in a simplistic manner, and your numbers are weakly established.
If people are backstabbing each other, they should be fired entirely separately from this policy. If you have no system for detecting and re
Re:So? (Score:4, Interesting)
Tip: When you're arguing with someone who actually worked under the system you're advocating, they're going to have much better insight into what happens than your hopes of what happens.
You're waving your hands in a simplistic manner, and your numbers are weakly established.
Says the guy advocating for firing an percentage of people every year, and setting that number to a wild-ass guess.
If people are backstabbing each other, they should be fired entirely separately from this policy.
How? The entire reason it's a backstab is management doesn't know it happened.
And every employee is bashing every other employee in peer reviews to keep themselves above your arbitrary line, so backstabbers aren't going to stand out.
Re: (Score:3)
Yet you can't seem to see any particular evidence that stack ranking works long-term. You also can't manage to see all the companies that jumped on this bandwagon and then found it massively destructive. Nor can you manage to see any particular evidence of the rate at which you should fire people.
Stack ranking is the formula to create a company of sociopaths. The rate at which sociopaths who are very good at their job is not not higher than the general population. Which means when you commit to this, yo
Re:Unfortunately necessary (Score:5, Insightful)
At one time it was typical for certain companies to hire out of the top half of certain college graduating classes, train them, put them out in the field, and give them a year to prove themselves. If you are promoting from within, and building long term capacity, then stagnant workers are not useful. And paying some huge sums just because they have been sitting at a desk for 20 years is no longer feasible.
That sounds like age discrimination, but with more steps.
It's better than that. (Score:5, Interesting)
It's not just age discrimination, it's a good way for a company to repeat their past mistakes over and over again, because the institutional memory of past mistakes is lost. And worse, they don't have anyone with enough experience and integrity to tell management no when a plan clearly will not work.
I've seen firms blow millions of dollars on mistakes that an older, more experienced engineer would have avoided. I've also seen firms avoid million dollar mistakes because the required functionality was implemented with a registry fix rather than a 30,000 hour effort to build a simulator.
Re: (Score:2)
Amazon gets rather attached to their own special vocabulary. For software development engineers, Amazon had(has?) some goofy list of qualities that they look for like "fungible". Essentially an application, working with a recruiter, needs to explain in what way she is fungible.
Maybe Amazon hopes some employees so fungible they can show themselves to the door. But I doubt it, they let my coworker go and he wasn't even allowed to clean out his own desk. It was one of those scenarios where the time ran out on
Re: (Score:2)
They want mediocre programmers with very little specialized knowledge?
Re: (Score:2)
Most places want this. Us code monkeys do most of the work. You only need a few specialized engineers writing patents and designing architecture, the rest of us are just here to execute on their plans. And we're replaceable cogs in the system.
If a developer is not a fellow, distinguished, or principal engineer then they're probably in the code monkey category like myself.
Re:"unregretted attrition" (Score:4, Funny)
They want mediocre programmers with very little specialized knowledge?
No, they want people who can follow best practices and document their code.
Mediocre programmers with very little specialized knowledge are much less fungible than experts; they don't use known, named algorithms, they don't follow best practices, they don't do a good job at documentation, they don't even know how variables are expected to be named.
You can test for knowledge of these matters by asking people in what way they are fungible.
Re:Late stage capitalism at its finest. (Score:4, Insightful)
I really need to find alternatives to Amazon Prime and AWS.
LOL they don't care about your crying, the alternatives are more expensive and have less features and that's why you're their customer.
For example, you can shop at newegg, and the products actually sold by newegg might be better, but their "marketplace" sellers might actually be worse. And you'll usually pay more.
With AWS... you'll pay more for less features are worse uptime.
Re: (Score:3)
This always sounds like something that would happen to circle-jerk armchair managers, but does Amazon actually have a hard time hiring?
The examples were it ends poorly are usually places like MS-under-Ballmer, where it was combined with general mismanagement and tolerated hostility between teams. In that situation, the problem is that the company tries to fire under-performers, but fires somebody else, because their metrics are all baked.
Google is still a highly rated place to work, they still hire up a lot
Re: (Score:3)
Microsoft didn't make the Fortune list either, though you will find them in the top 50 in some surveys. Still a toxic cesspool, but Google, always having a bad case of Microsoft envy, finally did supercede them in that particular ignoble way.
Re: (Score:3)
The Forbes list has Microsoft at #2 for Washington State, only Costco is higher.
They beat Apple, Delta Air Lines, Google, Starbucks, REI, AT&T, even WSU.
You could have at least checked the list before making the claim.
Re: (Score:3)
Unfortunately, most first level technical managers are horrible at firing people and are hesitant to do so with workers that are accomplishing something useful even if the net output doesn't justify the expense. They are also hesitant to do so because (1) they now may have an additional opening to fill which is rarely their favorite task and (2) they might lose the opening on the next belt tightening if there isn't a human in the slot and they will then have to go to the hassle of getting a new opening allo