The problem with stack ranking at places like microsoft was twofold. First small team managers were forced to rank their reports, and each manager had to get rid of the bottom 10% each year. If you had 20 employees you had to fire 2 employees each year. But statistics doesn't work like that, there were many microsoft teams with all productive employees. And there were probably teams where 30% of the group should have been moved/retrained/let go.
The second is that people respond to incentives. So a manager might make a bad hire because so they don't have to fire any of the people they like. The team might not help someone struggling because they need a sacrifice every year. And finally people might trade to barely avoid the bottom 10% (if you'll make this decision i want, i'll rank you higher). Then the people you are keeping are the low productivity, agile political operators. And a low productivity highly political worker is 10 times worse for productivity than the plain version of low productivity worker.
There doesn't seem to be any indication that Amazon is forcing this kind of decision on small teams. On a larger scale of 1,000 or 10,000 workers it makes sense to track if you are losing people that you don't care about losing or "un-regretted attrition". It's also good to have a process where a manager says "this person isn't doing well" and HR says "tell them what they need to improve and give them a few months" (or find them a different position). And the manager says "hey that worked" 2/3 of the time and then fires them when it doesn't. Otherwise you end up with sudden surprise firings or managers who never fire anyone.
It did not work like that at Microsoft at all. Stack ranking happened at an organization-wide level, so the total pool was hundreds of employees. It was perfectly normal for one team to be full of high performers, and vice versa. And being ranked in the bottom bucket didn't mean you were going to be fired. In most cases it wouldn't even lead to a PIP. It was simply a lower bonus, that's it. I have no idea about Amazon but at Microsoft the whole manager having to fire 1 out of 5 employees every year was 100% a myth. I worked there for many years and out of many hundreds of coworkers maybe 2 or 3 were ever fired for underperforming (while it should have been way more).
Funny enough the move away from stack ranking was very unpopular at Microsoft, since the performance review process became very opaque and way more political.
I ended up being on the losing end of the stack ranking at Microsoft and it was pretty terrible.
My manager straight up told me that my performance was fine but because he had to rank someone the lowest, it was my turn since I was the least experienced on the team. If I had actually been doing a bad job it wouldn't have felt as bad because at least then I would have had some control over the situation.
I moved teams shortly after that and my new manager was pretty surprised to see my last review. He started keeping a list of other developers on the other teams that I was helping so he could help me get ranked higher. That felt super shitty and I left the company right after.
That happened to be right at the same time they did away with the stack ranking. I have no idea if it improved things but it had to be better than what it was. (At least in my part of the company.)
The current company I work at uses stack ranking. I had no idea what it was until I joined a new team and was told, "Well, no bonus for you, because you're the new guy. Mark always uses the new devs to rank against so he can give out bigger bonuses to his cronies."
Sure as shit, for three years, it didn't matter how awesome or shitty a job I did, I got the same middle of the road review. One year, I got a paltry bonus and no salary increase. The next was neither and the final year I got a pittance of both. By the second year, I knew what was happening and just completely checked out. I went full "Office Space" and did the exact minimum in order to get my work done and not get put on a pip. Would come in around 11, get my stuff done in a few hours and tell my team I was going to be "WFH" the rest of the day.
Finally left that team and two years since leaving that team, I've had several large bonuses and salary increases.
Stack ranking is some toxic bullshit and managers use it to make themselves look better to their own skip managers. Fucking corporate politics man, it ruins everything.
That's also the case at Amazon. And likely most large organizations. It's not a strict stack rank but an expectation that over a certain population size, there is a certain percentage of under-performers. It's also used as a method of calibration. Performance evaluations are hard.
In my experience at Amazon, line managers (5-20 people) still often act as if they have unregretted attrition goals.
It's up to the Sr managers and directors if they spread the 6% goal around all their teams, or let one or two teams "implode". But the line managers know that their team either needs to outperform other teams or that the performance management buzzsaw is likely coming for one or more of their reports.
Heck, I even saw directors and VP refer to "regretted attrition" and "unregretted attrition" in all hands meetings.
My impression is that most teams have high enough natural attrition that it doesn't really matter and some teams even seem completely immune to any sort of real change, even if they do extremely poorly on their deliverables. I know a few teams that other teams actively avoid working with, even preferring to reinvent some of their solutions, yet, more or less the same people have been working on these underperforming teams for several years.
I’ve noticed this too in my time in my org, it’s reasonably common to just rebuild something simple if the other team isn’t responding to your tickets and escalations. The fact that they are three or four different time zones away from me on either side of the country doesn’t help.
I get that Amazon is known for their cold, calculated workforce, but it really takes a special kind of person to execute a, "hire someone just to fire them no matter how they do" plan over the course of like, an entire year, and honestly it's not even in that manager's best interest if the new person is actually pretty good.
Are we really suggesting that's the norm at Amazon? Maybe, but wow.
The link you pointed to mentions how hire managers are looking to hiring as many good people as possible. Of course they are if so many people are being fired.
It says that managers who don't perform well will be out. And that was the reason given for why this practice doesn't happen. Wouldn't the existing employees know the system better and be more valuable? Unless you want to get rid of someone it is better to keep your existing staff. Team members can't help the new person because them doing well would mean putting yourself at risk. It is setup to be toxic.
> Of course they are if so many people are being fired.
You don't seem to understand. Amazon grows by 15-30% every single year.
We're constantly hiring, even if there is no attrition. We also have a high bar. The ~ ratio is 20 phone screens to 4 onsites to 1 hire. That's a lot of work to get a butt in a seat.
> Unless you want to get rid of someone it is better to keep your existing staff.
Only if you presume that every single team has to identify that X% of people for the URA target. They don't. Period. Full Stop. It's a lie.
> Team members can't help the new person because them doing well would mean putting yourself at risk
The #1 thing that gets asked about individuals during yearly operational reviews is "what did they deliver"? We try to identify individual contributions but individuals can't ship. Teams ship.
You can still have bad hires. It happens rarely that the bar raising process outright fails, but it still happens, since the data you get out of an interview is very limited. For example, we had a team member that not only was doing a bad job on his own stuff, he was a drain on morale and resources for the whole team, since he constantly complained about everything, he never tried to learn anything, he posted big commit dumps every few months - hundreds of commits at once - and the code quality on those was abysmal. Luckily we got rid of him pretty fast, but not fast enough.
Most bar raisers are limp noodles that will just do whatever the hiring manager says. It's trivially easy to hire a mediocre person with full intent to fire them, and a bar raiser would likely be none-the-wiser nor would they stop it.
Companies implement stack ranking because the management does not have necessarily visibility to valuate each employee. Those who can't stand stack ranking should either join a small startup, where visibility is not issue by nature, or join a team with explosive growth, so each person's productivity and impact can easily be gauged.
No, I don't think stack ranking is necessary. It, however, could be a consequence of bad management or of the belief that stack ranking is the most effective way of continuously improving teams.
>> could be a consequence of bad management or of the belief that stack ranking is the most effective way of continuously improving teams.
I've never been at a company who uses it where the teams improved at all. Most of the time those regressed since people were actively working in silo's and shitting on other people's work to make their own work appear better.
Google's perf tools used to support stack ranking and layering. I used to love it...because it was completely open ended. You could stack rank on any dimension you wanted to, and you didn't even have to say what the dimension was. You could stack rack anybody, over arbitrary sets. AFAIK the results were only ever used in aggregate -- without the dimensions there really wasn't any other way to use it.
You could stack rank how nice you thought people were, making two layers -- like "really nice" and "mostly nice" (of course, there were no labels in the tool). Or you could use it to highlight two particular team-mates who really stood out by making a layer for them, and then a layer for everyone else.
It could also be used to stack-rank the entire management chain, which was fun.
The second is that people respond to incentives. So a manager might make a bad hire because so they don't have to fire any of the people they like. The team might not help someone struggling because they need a sacrifice every year. And finally people might trade to barely avoid the bottom 10% (if you'll make this decision i want, i'll rank you higher). Then the people you are keeping are the low productivity, agile political operators. And a low productivity highly political worker is 10 times worse for productivity than the plain version of low productivity worker.
There doesn't seem to be any indication that Amazon is forcing this kind of decision on small teams. On a larger scale of 1,000 or 10,000 workers it makes sense to track if you are losing people that you don't care about losing or "un-regretted attrition". It's also good to have a process where a manager says "this person isn't doing well" and HR says "tell them what they need to improve and give them a few months" (or find them a different position). And the manager says "hey that worked" 2/3 of the time and then fires them when it doesn't. Otherwise you end up with sudden surprise firings or managers who never fire anyone.