Really? You must be on a rare and exceptional org, then. I left Microsoft because the stack rank system most often worked exactly as described in the article and it was just as corrosive as they described. (I’m making more money and having way more fun now, too.)
In my extensive experience as both an IC and a lead at Microsoft the curve is enforced down to at least the dev manager level (i.e. teams of 20-30 people). A manager with 20+ people often ends up having to pick two sacrificial victims to receive a punitive review score (i.e. a 5 in the current system) even if everyone’s performing well. This causes all kinds of unintended side-effects and has absolutely rotted the company from the inside out.
I understand that the idea is to keep everyone on their toes and encourage them to keep pushing. It might even work ok if there were some fool-proof way to guarantee that everyone always got evaluated on their actual product work, but engineers are smart people and they usually figure out that it’s easier and faster to optimize for networking/schmoozing/backstabbing than it is to merely do brilliant engineering.
If a low review score was just a momentary, “better luck next time” sort of thing then it still might work ok but a low review score is a terribly punitive thing. You get no raise/bonus/stock, you’re not allowed to interview with other teams, and you’re in serious danger of being fired. It drastically affects your career prospects for the next several years. If that happens to you, it’s not worth sticking around. Because a low review score is so punitive, people are highly motivated to do anything possible to avoid it. The most efficient and effective means of avoid it is often not to spend time doing great engineering, but rather to spend time on non-engineering tasks that “increase your visibility”, or doing things that make everyone else look bad. Some people are quite good at doing that and they tend to be rewarded for it.
Sometimes low review scores are handed out to people who deserve them but often they go to people who just got caught at the wrong end of the political network. Being “the new guy” on a team as a result of a reorg is one of the most hazardous places to be at Microsoft. If a manager needs to dump a low review score on someone, he’s going to choose the most expendable (i.e. least well-known and well-connected) person. The point is that you can easily end up with a low review score through absolutely no fault of your own, then your career at Microsoft is screwed. That kind of chronic stress and fear shapes the company and has made it what it is today.
Granted, not every team at Microsoft is completely disfunctional, but enough of them are that the company tends to bleed great engineers and visionaries. The proof is in the stock price.
In my extensive experience as both an IC and a lead at Microsoft the curve is enforced down to at least the dev manager level (i.e. teams of 20-30 people). A manager with 20+ people often ends up having to pick two sacrificial victims to receive a punitive review score (i.e. a 5 in the current system) even if everyone’s performing well. This causes all kinds of unintended side-effects and has absolutely rotted the company from the inside out.
I understand that the idea is to keep everyone on their toes and encourage them to keep pushing. It might even work ok if there were some fool-proof way to guarantee that everyone always got evaluated on their actual product work, but engineers are smart people and they usually figure out that it’s easier and faster to optimize for networking/schmoozing/backstabbing than it is to merely do brilliant engineering.
If a low review score was just a momentary, “better luck next time” sort of thing then it still might work ok but a low review score is a terribly punitive thing. You get no raise/bonus/stock, you’re not allowed to interview with other teams, and you’re in serious danger of being fired. It drastically affects your career prospects for the next several years. If that happens to you, it’s not worth sticking around. Because a low review score is so punitive, people are highly motivated to do anything possible to avoid it. The most efficient and effective means of avoid it is often not to spend time doing great engineering, but rather to spend time on non-engineering tasks that “increase your visibility”, or doing things that make everyone else look bad. Some people are quite good at doing that and they tend to be rewarded for it.
Sometimes low review scores are handed out to people who deserve them but often they go to people who just got caught at the wrong end of the political network. Being “the new guy” on a team as a result of a reorg is one of the most hazardous places to be at Microsoft. If a manager needs to dump a low review score on someone, he’s going to choose the most expendable (i.e. least well-known and well-connected) person. The point is that you can easily end up with a low review score through absolutely no fault of your own, then your career at Microsoft is screwed. That kind of chronic stress and fear shapes the company and has made it what it is today.
Granted, not every team at Microsoft is completely disfunctional, but enough of them are that the company tends to bleed great engineers and visionaries. The proof is in the stock price.