Former Amazon SDM here. The actual rating process didn't strike me as something all that controversial. I had to do extensive write ups for all of my reports--this was standard stuff: what went well ("superpowers"), what didn't ("growth areas"), and some actionable ideas for improvement.
Then I had to pick numerical scores for "impact" and "potential" that were fed into a black box that resulted in an "overall value" score--this score is meant to be mysterious to the employees but it's the worst kept secret in Amazon at this point. OV ranged from LE (least effort/effective) at the low end, through HV 1,2,3 (high value) which is essentially "meets", and then TT ("top tier") at the high end.
~10-15% of a director org is supposed to be LE, up to 20% TT, and the rest of people distributed through the HV levels. I proposed these ratings to the bigger org that were then agreed to or adjusted by HR and higher management in a formal manager meeting.
The infamous URA quota of ~6% unregretted attrition (PIP) is meant to be met through managing out LE employees. The process starts with a PIP-lite called "Focus" and goes to the real PIP at the next stage, called "Pivot".
Many people got out of Focus, very few get out of Pivot. The part of the process I had trouble with, which contributed to me leaving, is the curve fitting of ratings for people. I had to do things like trade a HV3 down to an HV2 in order to keep an unpopular but otherwise fine performing engineer from getting an LE.
The stories people tell about hire to fire and unfair Pivot are overblown in my experience but not false. From what I've seen, the company will find pliant people and lean on them hard for more and more and more and then manage them out when then start to falter.
Lot of nuance to this--I have no love for Amazon and I'm happy to have moved on. It is definitely a meat grinder. That said articles like this are pretty low effort on my view.
Everything here is true. And the article is really low effort.
LE and URA targets do change by organization and current HC goals. People need to go to Focus but in my experience it was the right thing to do. HR will absolutely help you to improve the performance and retain the individual if you think it's fixable.
From my experience, and I'm not saying this is case for all, people in Focus had undeniable performance problems that needed addressing. Stack ranking and quotas suck, and you feel like shit having to abide by them, but I've also seen managers being overprotective about underperforming employees. At some point, you have to do something about them because they'll bring the whole team down.
The problem with zero sum game performance reviews is that they're not about improving performance but about fitting everyone in the same boxes. And this usually results in ridiculous outcomes, some of which are particularly bad depending on how that "box fitting" is implemented.
> trade a HV3 down to an HV2 in order to keep an unpopular but otherwise fine performing engineer from getting an LE.
Like this where the company's goal is never to improve "unpopular but otherwise fine performing engineer" but rather to keep the boundary of the boxes. And this despite mounting evidence against such systems, and the known abuse with methods like hire to fire. To me this is always a sign of rotten leadership.
Can't add to Amazon, but wanted to add another example of stack ranking problems.
I and some friends worked for Capital One, which does stack ranking (but not the intense PIP that Amazon does).
Overall the stack ranking isn't a huge issue, but you get screwed because those boxes are subjective, which I always felt was the big issue with any system like this.
I worked easily 9-6 most days and got an average rating, while I buddy of mine in a different org would work maybe 4 hours a day and get the same rating. My org had intense work and my manager saw that as the norm.
Again not terrible, but I would be incredibly mad if say this same system conspired to putting me on PIP. Working more and longer hours but not meeting the definition of average for your specific manager and team.
> Stack ranking and quotas suck, and you feel like shit having to abide by them, but I've also seen managers being overprotective about underperforming employees.
Then those managers are bad at their jobs and should be PIP-ed even more ruthlessly.
It's always interesting to me how managers get a complete and utterly unexamined pass in these conversations. The bar is so, so low for them. Managerial incompetence can be breezily cited as the reason for a horrible policy, like managing out a fixed percentage of every team, and everyone considers it a fait accompli. After all, what would be the alternative? Managers actually being held accountable for doing the most important part of their jobs?
ICs must be held nose to the grindstone every minute of every day... but ya can't do anything about bad management!
Umm, no... directors & VPs do get ruthlessly Lord of the Flies'ed (new verb here).
Middle managers, other than maybe first-level ones, are the lowest of the low, and very rarely get managed out. In fact, they are the cause of nearly all corporate problems, IMHO.
I've seen one incredibly horrible manager managed out this way. But anyone above incredibly horrible tends to survive all the reorgs.
If the reorgs actually got rid of bad management, it wouldn't be necessary to shrug and cite bad management as the excuse for mandatory stack ranking and attrition quotas.
"We can't help it! Our managers can't be trusted to tell us if someone is effective!"
Being a (low level) manager sucks, though. At least if you're a competent programmer and could get by with a very similar pay with little involvement in politics.
Another former amazon manager here, I can confirm this. The rule of thumb was %6 URA target and a somewhat larger number into Focus.
I never saw any written documentation about the URA goal, HR probably wants to avoid creating a leakable paper trail. It always came down verbally from my L8. I heard rumors that Amazon's HR boss Galetti is the big advocate of the URA goal, but I was nowhere high enough to be part of those conversations.
Only one nitpick - I remember LE as being around 10% of the org and was strongly suggested to put around 15% on focus so that 6% can be managed out, assuming 50% of people on focus would get out of it.
Yes, N level managers manage ICs and managers to N level. There are occasional strange cases where there is a L5 manager with a L6 IC report because of a weird org structure (small engineering team assigned to a non-tech org), but that isn't normal.
If you're in the industry and used to this, it "doesn't seem crazy" but it feels completely insane to most other professions. I only realized how completely lord of the flies it was after reading this and thinking about it if I was a teacher, doctor, lawyer, firefighter, journalist...
I can absolutely understand how this feels news worthy to those outside of tech.
It should be more controversial and the fact that people are saying "hey this is just how it works" is part of the problem.
I'm not sure where you're getting that impression from, law (and many other professions that operate under partnership like structures) often have an up or out system that is significantly more brutal than this. No room for expert operators to stay at a particular level, you get promoted or you leave.
A friend’s father who was a manager at a giant US conglomerate in the 1960’s and 1970’s did have this to say about stack ranking — they were more willing to take a chance hiring someone from a non traditional background or with a weak resume but good interpersonal skills. Often times it wouldn’t work out but occasionally they might find a diamond in the rough, but they’d be willing to take the chance because they knew they’d be getting rid of people anyway.
I think the main issue is to the more common person that paragraph read like as bunch of complete business nonsense. I have no idea what the vast majority of those abbreviations mean.
For Focus and Pivot do the engineers need to sign paperwork saying they are acknowledging being put into the URA threshold/PIP program?
What about if someone leaves and they were never officially put into Focus or Pivot? I've heard of someone leaving where their manager never put them through any part of PIP, and then after they left they found out they weren't allowed to rejoin Amazon for a year. IIRC Amazon lets employees rejoin up to 6 months after leaving, maybe longer, usually.
I believe so, at least through L7 (senior manager). I was L6 and only did a few manager calibrations before I left. Another commenter said they were an L7 SDM, they would know better than I. The sense I had was that things work differently for L8 (director/senior principal) and above but I have no first hand experience or knowledge of that.
Hire for fire feels like an urban myth. We do a lot of work to find you, interview you, hire you, and train you. Managing someone out is also a lot of work.
That being said a company as big as Amazon will always have edge cases.
The myth I heard was:
Manager has 10 employees who are absolute rock-star engineers. They love their team, work great together, and deliver amazing results.
This manager obviously doesn't want to hurt this awesome team or lose any members, but Amazon wants them to churn out the bad engineers (which this team may not have).
So instead of finding faults where they don't exist, our manager hires 2 engineers with little to no chance of working out in the team long term, give them some busy work, puts them on a PIP, then eventually fires them, only to soon replace them with two new victims.
It sucks to be a victim, and it sucks to knowingly do that as the manager, but it does work out great for the team and the manager - so I think it could happen (maybe not super common, but it isn't unrealistic in my mind).
That's the problem with applying a curve as a rule. In a large org it's probably true across the board. But there will be outlier teams. How does an org handle those?
You can critique Amazon for basically saying "don't worry about it" so managers may invent plans like hire-to-fire to game the system and overall someone's gonna get screwed on that team, but it's a tough problem to solve - otherwise, what's to stop managers from all simply claiming "no, my team's a special case"? Introduce special cases and it'll just be gamed through that mechanism.
I'm a firm believer that if you want to avoid things like that you have to avoid large organizations. A large organization is highly incentivized to be bureaucratic so to minimize the effects of unexpected losses and to keep the money machine running.
>otherwise, what's to stop managers from all simply claiming "no, my team's a special case"? Introduce special cases and it'll just be gamed through that mechanism.
You lost me. Why should we want to stop managers from saying that?
You got the myth part right: having a single team with 10 rock-star engineers.
1. Amazon isn't so special that even they can control the distribution curve to only have rockstars,
2. This is too big for a single team in the first place, let alone 10 rockstars,
3. You don't want any rockstars, let alone only rockstars on your teams,
4. Managing low performers is a magnitude more work than solid performers; no
manager would do this to themselves on purpose. It would be easier to cut their 2 least rockin' stars.
I hate that some people say rockstars and mean ‘really great developers that write beautiful working well documented code quickly and work great as part of a team’.
And other people say rockstars and mean ‘assholes that churn out new features really fast with incomprehensible code and then leave others to maintain their monstrosity as they move on to the next thing”.
I assume Amazon is large enough to have a team somewhere with 6-10 of the ‘good’ version, and also a team somewhere with 6-10 of the assholes.
> Managing low performers is a magnitude more work than solid performers; no manager would do this to themselves on purpose. It would be easier to cut their 2 least rockin' stars.
Then what do you do a year (and a half for Amazon) later for the next cycle?
One of the many problems with stack rank cull N% of your employees every year is that "every year" part. At whatever level it's mindlessly applied, if the company wants to keep it at the same head count you by definition have a steady churn of hires, which I'll grant you won't always be good, and fires.
> Hire for fire feels like an urban myth. We do a lot of work to find you, interview you, hire you, and train you
Your unspoken assumption here is that recruiters, HR and managers have perfectly aligned incentives. It is trivial to find instances when this is not the case - e.g. a high-performance team being forced to fire their "least effective" member[s]- this relies on the theory that talent is uniformly distributed across Amazon.
Amazon manager here: I just did a mental count of our URAs while I have been at Amazon. More than 3/4 are new hires fired in less than a year.
Hire-to-fire is neither a conspiracy nor a conscious tactic. It's the natural consequence of what we managers are put into.
If HR wants you to URA someone out, who would you rather lose? The experienced SDE who is productive and is responsible for projects you don't want delays in; or the new hire who's still trying to learn company tools.
I wouldn't say it's a "usual" thing at all, but more common the lower your level. L4 or SDE 1 has highest risk for negative personnel actions in general for 2 reasons: there are a ton of people in this level and many of them are early career and don't know how to protect themselves from corporate politics.
You don't have a lot of control over that if you're not a manager. Early career politics is basically your relationship with your coworkers and your manager.
Ideally you'll have coworkers that want to succeed and want you to succeed so that you are all open to learning from each other.
Sometimes you have a coworker who has gaps in their knowledge and is insecure (or sometimes that person is you). To avoid politics there, just help that person feel secure -- find things to learn from them or find ways they can help you.
Maybe a coworker is burnt out and doesn't have the energy to work as hard as you can. Learn what you can from them too, help them feel important.
Maybe your manager is overloaded with work. Figure out what their responsibilities are so you can prioritize what you work on in a way that makes their life easier.
Basically: politics is inevitable but you should just be new, make friends, have empathy, and let things fall into place.
This advice to avoid politics by becoming a people pleaser is not a bad strategy to attain the goal, but it might come at the cost of your sanity. The only way to be not involved in politics is to totally accept the current situation and to align your goals and interests with those in power.
This is the type of behavior that causes mid-life crises.
> The only way to be not involved in politics is to totally accept the current situation and to align your goals and interests with those in power.
At a small, young, growing company, that's what the company needs - if your goals aren't aligned with the company's, you are going to be spending scarce resources on the wrong things.
If there's a way, in a larger company, to prevent "necessary alignment" from turning into "political gamesmanship," I've never seen it. If the people in control of your fate have goals that aren't perfectly aligned with "overall health of company" then now it's games and bullshit and it gets real hard on the individuals. And that can happen very easily: Director X wants to be promoted to VP, needs to show results in their domain, maybe doesn't care if this has knock-on effects on other orgs; the company is 99% likely to be fine anyway, and it'll help them move up... do you get on board, or do you push back because it's a bad overall strategy? Good luck... :|
>At a small, young, growing company, that's what the company needs - if your goals aren't aligned with the company's, you are going to be spending scarce resources on the wrong things.
This is extremely true. I should have also added that sometimes situations and people are just toxic, and when that happens all you can do is try to get out of those situations. If you have burnt out coworkers and an overworked manager -- these might also be red flags
Some people strongly disagree with playing the game this way: however parent is fantastic advice because you need to know how others are playing to win, and you need to recognise those players and fit your game style appropriately to adjust for that.
Any company with more than 100 people will display these behaviours.
Companies with > 1000 employees will operate completely like this. Where there will always be 3 “priorities” for every manager: what they were told by the C-level, what they know to be important to their manager and what _they_ want to get out of their position.
The amount of times I have been caught totally off guard because I couldn’t fathom why a certain team or individual would be actively doing something to the detriment of the business… only to later understand it was to the benefit of their unit/selves.
The biggest issue I've seen with new engineers is a mismatch between what behaviours they think the company should reward them for, vs the actual things the company judges performance on. Pay attention to some senior people who are doing well and see what it is that they're doing to be successful. Hint: it's usually not writing the best/most code.
Often what is considered success at L6 is different than at L4 so this should be done very carefully. Generally at lower levels writing good code is much more important.
I’ve seen new engineers that tried to rush senior eng behaviours and they didn’t do well in performance reviews until they corrected their behavior. You need to prove you can be a baseline contributor before you start dreaming big.
Levels are another thing that are management done poorly, like management by stack ranking. "Levels" are usually made up with little concern for actually creating value or helping employees be awesome. Rather, they serve as a formalism for forcing objects of various shapes and sizes through round holes because 0) they can't think of anything better and 1) other large companies do it, so it must be right.
Work on things that the firm values[1], for bosses who will reciprocate your work at making them look good, by making you look good.
Your manager has a huge amount of impact on your career. If you're working for an asshole/moron who won't reciprocate in this way, find a different one.
[1] Not the same thing as work that's valuable for the firm.
That's an enormous topic that is hard for me to treat here. Knowing that it's a thing you should do is a great first step though. Find more experienced people in your context willing to mentor you and critically listen to what they say.
Don’t go to companies with corporate politics. If you’re going to such a company (where politics are important, like Amazon) you must participate: you cannot protect yourself from them.
Any company develops corporate politics. Corporate politics stems from the corporation ultimately being a non-human entity that doesn't really have desires and needs, while its individual humans do have them.
At some point what's "good for the company" will not align with what certain (powerful) individuals want or need.
Flattery until you get past the BS peddler managers. Once you’re in cahoots with the senior leadership you’re good. Honestly low and mid level managers are the worst, most of them suck at programming even if they won’t admit it, and their only strength is saying a bunch of words most of which are bullshit. They are there to be scapegoats, and you can safely ignore them.
Once you’re among the senior leadership flattery won’t work and honestly they’ll see right through it. If you play your cards right you’re free to prove your technical prowess and get placed into top performing teams with the best managers. As long as the senior leadership knows your name, and you don’t overextend like a noob, you’ll get what you want pretty quick.
* Be the only one who understands a particular codebase
* Listen when people give you advice, so that they fee they have some hand in guiding you, perhaps even carefully encourage advice-giving
* Make some friends just outside your immediate team
* Avoid letting people get frustrated at you or your work.
* Gossip nice things about people, they'll hear about it
"I had to do things like trade a HV3 down to an HV2 in order to keep an unpopular but otherwise fine performing engineer from getting an LE."
And you did this because why, exactly? It doesn't matter if someone is performing fine if they're an insufferable ass to everyone around them and breeding resentment, etc.
Good summary. This is probably the most accurate and succinct writeup I've seen (in what is otherwise a sea of misinformation about the company). I'm also an ex Amazonian (Sr SDM).
Former Amazon SDM here. The actual rating process didn't strike me as something all that controversial. I had to do extensive write ups for all of my reports--this was standard stuff: what went well ("superpowers"), what didn't ("growth areas"), and some actionable ideas for improvement.
Then I had to pick numerical scores for "impact" and "potential" that were fed into a black box that resulted in an "overall value" score--this score is meant to be mysterious to the employees but it's the worst kept secret in Amazon at this point. OV ranged from LE (least effort/effective) at the low end, through HV 1,2,3 (high value) which is essentially "meets", and then TT ("top tier") at the high end.
~10-15% of a director org is supposed to be LE, up to 20% TT, and the rest of people distributed through the HV levels. I proposed these ratings to the bigger org that were then agreed to or adjusted by HR and higher management in a formal manager meeting.
The infamous URA quota of ~6% unregretted attrition (PIP) is meant to be met through managing out LE employees. The process starts with a PIP-lite called "Focus" and goes to the real PIP at the next stage, called "Pivot".
Many people got out of Focus, very few get out of Pivot. The part of the process I had trouble with, which contributed to me leaving, is the curve fitting of ratings for people. I had to do things like trade a HV3 down to an HV2 in order to keep an unpopular but otherwise fine performing engineer from getting an LE.
The stories people tell about hire to fire and unfair Pivot are overblown in my experience but not false. From what I've seen, the company will find pliant people and lean on them hard for more and more and more and then manage them out when then start to falter.
Lot of nuance to this--I have no love for Amazon and I'm happy to have moved on. It is definitely a meat grinder. That said articles like this are pretty low effort on my view.