Speakig as someone who is an interviewer, your point means nothing. The interview process is tailored to get people in, but it's totally independent of the decisions to decimate teams.
Where managers step in is in the company's yearly performance review process, and those tend to be brutal. unlike the hiring process, that has higher-ups (L7 and up) involved in deciding who performed according to the manager's expectations.
> How? It's not their choice.
The choice of who gets in is irrelevant. It's the choice of who gets out that matters. The hiring process gets bodies on desks, and afterwards managers decimate them.
Getting the orange badge is a feat, and I've been to meetings where orange badges were quite blunt in stating new hires they were in the company when blue badges get in, and they will stay in the company when said blue badges get out. This is not new to anyone working in the company. What's the average tenure at Amazon? Less than two years?
You contradict your own point above
> has higher-ups (L7 and up) involved in deciding who performed according to the manager's expectations.
These higher ups tend to have extremely well-tuned bullshit detectors and it would be extremely hard for a manager to sabotage an employee and get them categorized as underperforming.
It would reflect worse on the manager not the employee.
Orange/red badges are typically a big deal because it is very rare for people to last at Amazon past even 2-3 years, nevermind 5 years.
1. The company has grown rapidly, so many/most people seem fairly new. That doesn't mean that people that have been here a long time are leaving (although some obviously are). 2. The tool available to employees shows tenure of all employees, including fulfillment center workers, which are much less likely to be long-term.
Are some career ladder jumpers actually planning to leave anyway after 2-3 years so they can leverage compensation offers between FANG?
In most cases such people are put on a plan, ironically called "Coaching Plan" itself, and thereafter fired.
URA quota is an open secret at Amazon, and some employee appreciate the practice of hire to fire, as it saves their own employment.
That in turn ensures that objection to bad hire, who is going to be "coached" is non existent.
The article posted here is pure speculation.
...and having worked for other large companies, I've seen similar metrics being tracked. It doesn't mean any of them are hiring just to fire people.
Frankly, I've noticed a LOT of poorly written anti-Amazon articles around, and I'm not clear if there's some effort to undermine the company. Is Amazon a political organization?
There’s a mandate to PIP and the manager wants to protect their productive employees so they hire someone and set them up to fail.
I don’t think you’re going to get any clean documentation on this process - it’s simply a result of a system that stack ranks and mandates firing.
About 15 years ago, I was placed on a PIP for low performance at a large technical employer. I managed to climb out of my hole and shine, but it's a good lesson in humility and responding constructively to feedback.
Citation needed.
Leaked internal memo shows that if an org doesn’t meet the target this year, they have to make it up next year.
Sure, fire has not been found, but there is a ton of smoke on this.
You are in a subthread that branched off of this statement:
> I was personally told by a former manager who used to work at Amazon that he hired to fire. He hated that the company forced him to fire good people and left.
So the conversation veered away from the BI article, and we're now focusing on the relevancy of anecdotal evidence that was presented by another HN member. The fact that you needed this explained completes the picture for me with regards to all your other comments.
I am not disputing that the NYT revealed his full name, they absolutely did so that, against his wishes.
"Oh @sokoloff quit? He was OK but let's mark him as non-regrettable so we can hit our target. Where do you want to get lunch?"
In happy cases like this, and in cases where you actually have significant performance problems, URA isn't so bad. The more degenerate situations happen in circumstances like:
* The employee wants to internal transfer. Bad Manager X immediately adds Employee Y to the development list and says "Oh ho ho, you're not able to transfer because of your performance problems." For Bad Manager X, the apparent incentive is high - either the employee stays and you keep the resource or they leave and you get URA credit. Obviously this leads to pretty awful situations, but I saw it happen multiple times. This can also happen when the manager is trying to protect the employee - e.g. add them to DL but not actually fire them. This usually backfires.
* Senior management pushes down URA targets. Hey, you've got a 15 person group? That means you need to pick one person to let go. Satisfied with all of their performance and understaffed to begin with? I don't care. Don't you want to have another great performer like employee Z? You need to make room, so tell me who is the least impressive person on your team. This leads to really sociopathic management behaviors.
All of this would be OK to me, if it was done transparently and more objectively. But Amazon has thrown out all the paper trail and messaged to employees that they aren't being stack ranked, which is complete bullshit.
Pro tip if you're working at Amazon: if you get a PCS with no additional grant and no base increase, you're on the development list. Find another job, or do something amazing.
Allegedly, the only sure-fire way of knowing if you are on Focus/Dev list is if you ask your manager point-blank about it and they have to respond with the truth.
The idea that I would put an employee onto some kind of dev/PIP/watch list and then not tell them about this fact is so fundamentally foreign to me, but I believe it happens (because no one would make up such a thing).
They’re one of the largest corporations in the world; of course they’re a political organization.
>Instead, Amazon managers are hiring people they otherwise wouldn't, or shouldn't, just so they can later fire them to hit their goal.
And as OP is correctly pointing out, that can't happen, because the bar raiser can veto any hire they think is unqualified. Or at least it can't happen without the bar raiser colluding - and since the bar raiser is not the hiring manager (or even from the same part of the organization as the hiring manager) and they are not being evaluated on the UAR, they have no incentive for this.
If, as you suggest, managers were hiring _qualified_ people with the intent of firing them... I guess there's no way to prove that isn't happening but FWIW, I'm also at Amazon and have been part of a bunch of interviews and I've seen no evidence of that.
For a person to get placed on a team at Amazon, the manager must be making some representation that they need another employee. Amazon isn’t going to waste headcount by giving it to managers who don’t ask for it. This means managers have control over whether hires are joining their team. Therefore a manager can execute the plan of hiring to fire. It doesn’t matter whether they have the power to approve a specific candidate.
Obviously if you are forced to cull X% you are going to try to maintain 1+X% staffing level.
Whether or not this actually happens, I don’t know. Do I think that managers (or really anyone else) can use soft skills to exert some influence some of the time? Having participated in many such panels, I think it’s pretty doable despite what many people claim because of the discussion process that happens.
The need to get vacancies filled.
Where is this data?
In practice, the HM absolutely does make the decision. If the candidate is seriously "below the bar", then the BR might strongly assert that they will not be hired. But in reality, most BRs do not actually uphold a very high bar. The BR's main function is to act as a facilitator during the debriefs, but if the HR feels strongly one way or another, the BR will almost always back up the HM.
If an HM wants to hire someone just so they can fire them, all they have to do is find a mediocre candidate that a BR will not feel strongly about, and then during the debrief they just say that they know the needs of their team better than the BR does, and the BR will say "ok, I'll go along with your decision". It's as easy as that.
This is especially so on teams that use the same BR over and over - hiring teams can actually influence which BR they use for loops, so if they find one that consistently disagrees with the HM, they can simply just find a new BR that is more agreeable.
It doesn't matter who you hire for that 10th spot. It doesn't matter who makes the decision. You're getting a 10th person, of any caliber, to get rid of them. You'll make open recs and add more people to the team when you don't need more people, just so you can keep your staff.
As a hiring engineer, this is exactly the sort of optimization I would make if I were compelled to terminate a portion of my team regularly (really, as a principal, I'd just tell whoever was making the rule to go to hell). We'd constantly be overstaffed so I can let someone go without it hurting my team.
Rather than “lower the bar” an HM can just request 1 more headcount for SDE1 than they truly need.
1) hey look, we're linking to someone else who said this before us
2) and please note, there's the obligatory "appears--at least on the surface" statement, so if this comes back to bite us, we told you you shouldn't have believed us in the first place.
Here's your startup idea: find a way to hold journalists accountable even when they operate within the legal grey area of journalism, which is becoming prevalent and will only continue to grow in intensity and harmfulness.
PS: I get it, you hate Amazon. So do I. Try to think for yourself for a minute and let's engage in this conversation as if we were talking about your favorite high school teacher who just had their career destroyed because some dude needed to get a few extra pageviews.
Parent's comment is not fully correct, yet you are taking it at face value and not looking "too deep underneath the carpet", and in some ways are "pick[ing] up someone else's lie and expand[ing] on it without bothering the verify the fundamental thesis".
The fact of the matter is that the Amazon HMs who decide on firing often are the same people who decide on hiring. Amazon may desire for that not to be true, but in practice, it is. And GP's comment doesn't actually provide any evidence against that.
I've added more context here: https://news.ycombinator.com/item?id=27371455, but the gist is that Bar Raisers are supposed to be an independent hiring evaluator, but IME they often just serve as meeting facilitators that will just back up whatever decision the HM makes, which makes the HM the ultimate decider.
Saying "my statement may or may not be right, I'll let the general public decide" is lazy and shirking responsibility. If you have reason to believe that your statements are wrong, you should retract those statements or at least put a disclaimer on them. I think that's the bar you would hold Inc to, and you should hold yourself to it as well.
The Inc journalist did a lazy job of presenting facts [1], and my suspicion is that this is not by coincidence - as a journalist, you either do a stellar job of researching every last bit of a story, or you do the opposite - close an eye and report as vaguely as possible, mostly citing external sources but somehow making the story sound original so you get some credit as well. Both are safe approaches; the space in between is dangerous - you can't take someone else's shitty reporting and then check if it's true before publishing - that's how you get sued for libel. Fun fact: if you prepend "in my opinion, ..." you're 100% safe from any libel risk.
If we consider additional evidence that came to light through the BI article, I am still aligned with mabbo's view - the Amazon management seems to have taken a conscious step against the hire-to-fire scenario by separating the ownership of hiring and firing.
[1] Note: this topic originally linked to an Inc article
And so are you. As has been pointed out multiple times, the assertions made about BRs and HMs in this thread are flat out incorrect, and yet you are continuing to rely on them and even go as far as to tout them in other comments in other threads as well.
>close an eye and report as vaguely as possible, mostly citing external sources but somehow making the story sound original so you get some credit as well
This is exactly what you are doing throughout this thread by repeating the notion that Amazon "separat[es] the ownership of hiring and firing", which if you were willing to do any amount of research rather than just blindly trusting a random internet comment, you would find just isn't true.
You are holding yourself to a different bar than Inc, which is hypocritical. The point about removing Amazon emotional context here also applies to Inc, because it's clear you personally have a lot of emotional baggage related to journalists. So the question is: would you post these same disparaging comments, and still maintain the same "my statement may be incorrect but I will continue to repeat it anyway" if it was your favorite high school teacher you were disparaging, potentially ruining their career? I would certainly hope not, so I don't see any reason other than arrogance of why you would do so here.
> the assertions made about BRs and HMs in this thread are flat out incorrect
The assertions are that there is a hire to fire practice. Those assertions are based on 1) the evidence of an attrition goal, and 2) anecdotal stories of people who know people. My argument is that 1) is not the same as a hire to fire practice [1], and 2) is hard to verify - there are a lot of people with an axe to grind. What I would love to see is an analysis of people who started their job at Amazon and didn't reach their first 12 months. If that rate is any higher than at FAANG for comparable positions, then we have a smoking gun. How hard would it be to do this? Not very. The lack of this effort is telling. Also, I tend not to subscribe to every conspiracy theory, which admittedly is hard these days. But you should try it, it would calm you down a bit.
> just blindly trusting a random internet comment
Thank you for making so many assumptions about me. You're wrong. It was a slow day today, and I was really intrigued by this concept, so I dug a bit deeper and read a lot of posts on this topic. My takeaway from all this is that this is a bit like a political discussion: some people really passionately hate Amazon's recruiting practices, and some defend it. I certainly wouldn't say that everyone is aligned that BRs and HMs are colluding and that the system is not working as designed.
> You are holding yourself to a different bar than Inc, which is hypocritical
That's a ridiculous statement. Journalists have a code of ethics, whereas the general population doesn't. This is a fact, let's please not debate something as fundamentally verifiable like this. Not everyone can do everyone's job. My job is not to spend my whole day verifying facts of what some journalist published, but it's very much the journalist's job to do that for their pieces. We're not in the same position, nor should we be. Not that this fact matters at all since I stand behind what I said, but you're starting your argument right off the bat with a fundamental inaccuracy.
> you personally have a lot of emotional baggage related to journalists
That would be weird - I worked for many years for the NYT and I think they are a wonderful org. So... your instincts are wrong again. I really hope you don't rely on your instincts for a living.
> would you post these same disparaging comments
Totally nonsensical reasoning once again. My comments are not disparaging, so let's just end that discussion right there.
If anything, let's turn our attention a bit to your interesting HN career. You created an HN account called tyuqfromhere 8 hours ago, which you used exclusively to attack Amazon. And then this new account, tyuqfromthere, was created an hour ago, and... you guessed it, it's used exclusively to attack Amazon. Someone said elsewhere that there's a lot of anti-Amazon propaganda recently, and I sure as hell would love to know who you are and what your real motives are.
To be clear, Amazon can crash and burn for all that I care, but there's one thing I do like about Jeff Bezos - it's his last sentence in his blackmail response letter [2]: "I prefer to stand up, roll this log over, and see what crawls out." I doubt we'll ever find out what kind of a creature you are and what you're trying to accomplish with your HN trolling, but it sure as hell would be interesting.
[1] I am not defending the attrition goal strategy, but its presence is not a proof of abuse. I know that if someone presented me with the requirement to let 2 of my 30 member team go within the next 12 months, and I was given 2 new team members to try out, I would definitely give everyone a fair chance. Also, every other manager I've ever worked with would do the same. I find it hard to believe that somehow that all managers at Amazon are suddenly different from anyone I've ever worked with.
[2] https://medium.com/@jeffreypbezos/no-thank-you-mr-pecker-146...
And btw, when you call people names, be aware that this often says more about you than the person you're trying to insult. Seems like you have a particular aversion against people with startup ideas, which automatically means that you were never successful with any of your own - people who make startups work for them don't have as much of a chip on their shoulder. That's ok, startups are hard and most people don't win - I have a lot of respect for anyone who even gives it a try. But please use your experience to help other people as opposed to just being bitter about it and putting others down.
Startups are great, but their chances of success are higher when the founders know what they’re talking about.
It's unfortunate that switching the source made it less clear which author's work I was commenting on, but I think that you'll see that most HN readers who engaged with me agreed that Jason Aten's article was rather weak. Your childish name calling, on the other hand, will always be unwelcome here regardless of circumstances, and please consider changing your approach or just go back to where you were prior to recently creating this HN account.
[1] https://www.inc.com/jason-aten/amazons-controversial-hire-to...
[2] https://www.businessinsider.com/amazon-tracks-unregretted-at...
Then others pick it up by referencing the first news report. They no longer have to be specific, instead its just "news source X reports".
Unless you follow the chain back to the source, you don't see the actual confidence level. Sometimes you don't even get an accurate picture of the original source.
There is no requisite for the person hiring to be the one firing for the person hiring to hire a sacrificial lamb.
A rational actor could easily decide 'Hey, I like my team, if I hire this guy/gal who the firing manager definitely won't like[1] He/she'll be fired before anyone on my team.' Morality aside its the same as serving up a secondary 'bad' idea to your boss to give him/her the illusion of choice, and get him/her to go with the plan you want.
[0]>So the core thesis of the article is fundamentally flawed, since the decision maker for hiring is not the same person who decides on firing
[1]For personal or professional reasons.
It's not even a news piece, it's an opinion column. You're free to disagree with it. That's totally fair.
I'm not reporting on Amazon's internal hiring practices, I'm applying a lesson to what Insider reported. That's not a cop out - it serves an entirely different purpose. In this case, it's a lesson for managers. If you don't feel like I made that point, ok, but I think the core thesis of your critique is fundamentally flawed.
Step one: treat any article that quotes anonymous sources with deep skepticism. The sheer number of retractions and stories that turn out to be blatantly false (especially in the realm of politics) has become absurd in the past few years.
No startup needed, just some (un)common sense.
You are a manager. You need five people on your team.
You hire eight, because you are expected to lose three in the next two years.
If three of them leave, fine. If three of them are incompetent, great! (You get your URA and you have evidence your Bar Raiser is an idiot.)
If the interview loop worked perfectly, all eight are great. But you still have to boot three of them.
1) Inc. just published a bunch of BS, which would have taken no more than a quick call to a hiring manager to validate or invalidate. The fact that they didn't even try is telling. The cost of publishing a lie is exactly 0, and the potential benefit is going viral and making a lot of $$$ with associated advertising impressions and new subscribers.
2) There's a conspiracy happening at a scale of a few hundred thousand people, who all do highly immoral things but nobody is telling on each other. The BR has no incentive to partake in the conspiracy, because their track record will show a high attrition rate which is against their goals, but they still decide to violate their specific job directives. And again, not one BR, but since this is systemic behavior, it's the majority of BRs across all teams and locations.
The culture of Amazon does not see this as immoral, so why would they "tell on" each other? And among those who maybe do see it as immoral, the managers are incentivized and compensated for it, and the non-managers are threatened with PIPs and firing if they disagree with it.
>The BR has no incentive to partake in the conspiracy, because their track record will show a high attrition rate which is against their goals
BRs do not have any kind of "track record" related to attrition rate. After a BR participates in a hiring loop, that's the end of the story for them. A BR has no incentive to not participate in this so called "conspiracy".
But furthermore, you have a fundamental misunderstanding of how the BR program works. In practice, the BR is just another interviewer that helps facilitate the debrief, and usually just defaults to echoing whatever hire decision the HM makes. If an HM wanted to hire-to-fire, the "participation" of BRs isn't required at all.
Why would any HM even speak to them?
> The fact that they didn't even try is telling.
How do you even know that?
> There's a conspiracy happening at a scale of a few hundred thousand people
This sounds like a gross overestimate.
This is not an organized conspiracy. Individual HM would come to the same conclusion, which is it is to their own benefit to act perversely when it comes to hiring/firing. This isn't even restricted to Amazon. Time and time again, at companies with poorly thought out policies, managers end up doing really shitty things in order to meet goals.
> who all do highly immoral things but nobody is telling on each other.
And because it's not an organized conspiracy, how would they tell on each other. It's not like they're gonna go around tell each other that they're doing corrupt things.
And even if they were, that usually ends up creating a code of omerta. You see it all the time in organizations where they do terrible things. These places end up creating a us vs them mentality. Police is one example. Another example is how China runs things. They spread out the bad actions across everyone, making everyone complicit. This makes people less likely to rat each other out because now they're the ones doing it and internally, people justify their own decisions.
> The BR has no incentive to partake in the conspiracy, because their track record will show a high attrition rate which is against their goals, but they still decide to violate their specific job directives.
This is the only thing I agree with. It is unlikely that a BR would collude. They have no reason to.
Which is the fundamental argument here. All this talk about how bad orgs behave this or that way is not relevant considering that they were exceptionally thoughtful about how to separate the hiring decisioning from firing decisioning.
I think we can all agree that Amazon does a lot of shitty things, but the concept of introducing a BR deserves praise.
Collusion between BRs and HMs is not the fundamental argument here. That was just speculation by someone else trying to poke holes in the argument that HMs do not hire just to fire. The fundamental argument is that the policies lead to perverse actions by HMs.
> All this talk about how bad orgs behave this or that way is not relevant considering that they were exceptionally thoughtful about how to separate the hiring decisioning from firing decisioning.
As has been discussed in many other comments, the separation of concerns here does not prevent the perverse action of hiring more people with simply so that HMs have more options of who to fire.
> so that HMs have more options of who to fire
Totally true, I don't deny that. But that's not the same as hire to fire, in the context of how it's commonly understood and criticized: that the person that was hired will also be the same person who is going to get fired. My point is that while it's unfortunate that someone will get fired, it's not necessarily the same person who was last hired.
One way to interpret that is that you are literally hiring someone just to fire them.
Another way, the way I and some others are using it, is that they're hiring people *so that* they *have* people to fire. As in, they're not necessarily firing the people they hire, but that they need to hire so that they have options when they need to fire.
I don't think the latter is any better than the former because they're still going about hiring with the explicit intention of firing people. Maybe it's not the hire specifically, but when they have that intention, it still creates the same kind of toxic environment.
As an Amazon Bar Raiser, our community takes our unbiased position and ability to veto hires EXTREMELY seriously.
We hire for Amazon, not for the team.
If on the other hand they have a meaningful influence then does it really matter if it's someone else that gives the "official" okay instead of straight out of their mouth?
And with the process changes in the US (and EU), candidates are now given the option of picking the team they want to join (which can be different from the interview-HM's team).
Do you happen to know how frequently a given Bar Raiser will be on the same interview loop as a given manager? Are they load-balanced across the whole company, or siloed to their home division?
It doesn't mean that they can't hire people to fire them. They could very well still be hiring people just to fire them because they care more about the people they already have or just to give themselves more choices to fire.
I make sure that managers reporting to me always make the final decision on hiring (and firing) for their teams. That makes it more fair to hold them accountable for their team, because it's the team they chose.
Moving the final decision to someone else, like a "bar raiser," seems fraught with misaligned incentives.
Are bar raisers somehow held accountable for their hiring decisions?
The bar raiser can say "no" even if the HM says "yes," but I never heard of it going the other way (where hypothetically the HM would say "no" and the bar raiser "overrule" the HM.)
And in hundreds of interviews there, I never experienced a situation in which it was as black and white as the above would make it seem. There were times, of course, when there was a difference of opinion, but it's on the people in the room to debate and convince each other based on logic and data - a bar raiser is not simply going to walk in and put his/her foot down and say "no" to a team that is otherwise saying "yes" without pretty detailed discussion and reasoning.
(When push comes to shove, the bar raiser could overrule, but everyone would rather it not have to get to that point...)
I have a ~10% hire rate in over 500 industry interviews conducted. These were at large tech companies so they included 5+ people spending 3+ hours each.
It’s not worth the time to fabricate head count to begin hiring, then fire a qualified person. We assume Amazon is cutthroat and grades on delivery. Imagine doing this sham while doing your normal job.
Edit: formatting
Either you haven't been part of the panel in a while or you haven't worked for Amazon. BR's responsibility is not to make the sole decision but help the panel achieve consensus. This is very clear in the interview training.
To be pedantic, the HM makes the hire-or-not decision, the BR can veto a hire decision. BR cannot make a hire decision.
If you think that this effort is demonstrative of their ability to translate needs into working code and then deliver it, I have a bridge to sell you.
I don't think it would be difficult for an experienced hiring manager to identify someone who's well-rehearsed but otherwise incapable and then hire that person specifically in order to fire him later.
In any interview loop, there are 4-6 interviewers. One is the hiring manager or a substitute for them. One is the Bar Raiser, an experienced interviewer with extra training. And the rest are guys like me who took training on how to interview.
When we've all had our time with the candidate, we meet back up to discuss. And at the end, the Bar Raiser- not the HM- makes the decision.
> Amazon managers are hiring people they otherwise wouldn't, or shouldn't, just so they can later fire them to hit their goal
How? It's not their choice. An HM can voice their opinion that a given candidate is great, they love them, they want them, etc, but they can't overrule the BR, who is trained to watch for bad calls by HMs.
If the author wanted to criticize the concept of URA, I'm with them. I don't think it's a good way to operate at all. But when you double down your criticism with untrue things in order to get better press, you undermine the original valid point.