Have you ever worked at or with MS? If so, you'd know that stack ranking is the reason MS has flopped in just about every field it's entered for the last decade.
As someone that works at Microsoft I know I can't really respond to this without being accused of being a shill/apoligist (even at a place that prides itself in logical thinking, like HN), but this is such a massive oversimplification, and likely so far off the mark as to possibly be on the wrong planet.
Microsoft has a well-known review system that engenders various feelings in various people, not all of them postive. I have posted here before how the stack ranking part is actually low on my list of objectionable things about the whole process, but it is what people know so it is what people cite. I have heard stories, always apocryphal, never by anyone willing to put their real name behind it, about how people are constantly obsessed with reviews and stabbing their teammates in the back left and right to get a leg up. I have never once witnessed that behavior or even heard of a remotely plausible instance of that happening. That said Microsoft is a big company, so I am sure it has/does happen somewhere. I dispute it is rampant or common.
I have heard stories of people that are terrible at working in groups and don't do well. I have heard stories of people that have an opinion of their own skills that seems objectively divorced from reality. I have heard stories of management hubris that ultimately doesn’t pan out. Some of the people that play a starring role in these dramas then invent conspiracies about plotting colleagues causing them to rank poorly, incompetent managers that didn’t recognize their delicate snow-flakism or any other number of explanations that conveniently don’t attribute any share of the blame to themselves. I suspect those tend to be fantasies made to soothe one’s ego.
I think anyone with knowledge of internal state at Microsoft could posit a much better explanation of why they have 'flopped in just about every field it's entered for the last decade' that doesn't involve Machiavellian competition between employees. Hell, reversion to the mean pretty much would cover most bases without any kind of comical soap opera drama needed.
I hadn't seen your previous reply, I don't always follow up on my own posts :) I wouldn't argue there is subjectivity in reward, there always will be. I don't work in Windows but I know a few folks that do. I haven't heard any stories of people feeling they got shafted and some doofus got a reward, but I am sure it happens, and I am sure it is not unique to Microsoft in any way. My argument is with people that claim it is the norm and they base that on their own (non-verifiable) experience, or worse on some troll blog like miniMSFT.
As for comments about seeing who got rewards based on title change, a title change doesn't always mean a large pay raise, and there are hidden rewards like gold star bonuses and HiPo groups that have little/no outward evidence, unless the recipient starts talking about them (which they tell them not to do), so I would say you probably aren't seeing accurate signaling based on job titles alone. The fact that some of the best rewards are secret is one of my beefs with the system.
Also, to clarify, I am not a fan of Microsoft's system, but when I hear stories of people who claim it leads to internecine wars, well that just isn't my experience in the last 8 years, or the experience of anyone I have talked to personally.
Ultimately there has to be some kind of ranking system as there is a fixed pool of rewards. The only alternative I could see would be exactly even sharing amongst a team. You could argue for that, but even as someone that has socialist sympathies I don't think it would work or be fair. There are people on my team that produce more than me, whether it be through being smarter, working longer hours, whatever. To say I should get the same reward as them is just wrong.
So if we agree there has to be ranking, well there's your stack ranking. You could argue with the curve fitting, but your claim that 1 in a group of 10 would get no bonus is just false. Curve fitting doesn't happen on a per team basis. It's more like 10 in a group of 100 or 100 in a group of 1000. There is nothing saying that an entire team couldn't end up in the top 10% of rewards. It may well be the case that on your team of 10 superstars one doesn't get the reward they feel they deserve, but that is because someone else in the calibration meeting was deemed more worthy, your teams ranking isn't done in isolation and in a team with 10 superstars there can be other superstars in the general vicinity, even more than 10. To claim that the system results in people plotting against the ones they are ranked against in order to sabotage them just sounds like conspiracy theory thinking. In my experience it results in people upping their game, or leaving if they feel they are unappreciated. Both are reasonable responses, some attempt at sabotage is not.
Easy for you to say. You're not being called apocryphal.
I thought readers could connect the dots but maybe a bit more explanation is due.
Having worked at the "new Microsoft" in recent history I feel like the part I quoted is highly indicative of someone who has their head in the sand. Across multiple divisions among people I've voiced opinions with openly I haven't talked to anyone at MS who shares his view. Even in his own DevDiv, I've heard people say that DevDiv is a bit better than the Microsoft average but yeah, it still has these problems.
At least, this is how I felt last month when replying to his previous comment. So I left a reply, even though it's taboo to talk this way about a former employer, even though I risked potential embarrassment attaching my name to this. And I said, hey, what they say about MS, what they said in that Vanity Fair article, it has a lot of truth in my experience, and I don't really know anyone who says otherwise without being a blind shill or a sociopath. (I'm saying it a bit more bluntly here but that was the idea.) I was reluctant to make that comment, but I did.
Then he goes around repeating still that all such comments are apocryphal and anonymous, despite the comment of a real person who is not writing anonymously. (Aside: He is probably not aware that current employees won't say this for fear of losing their job, and former employees probably feel the general taboo against speaking ill of a former employer.)
> Easy for you to say. You're not being called apocryphal.
You are literally just making my point over and over again. This sentence is, like the previous post, utterly devoid of logical content. Are you seriously taking offense at being called "apocryphal"? Nevermind that ryanmolden never actually called you apocryphal.
True that it may not be the reason they have flopped, but honestly every co-worker of mine that has worked at Microsoft states our review process drives performance above all else. And the sample size isn't small, I'm in SLU at Amazon.
Would you mind expanding on this a bit? How exactly does it contribute to Microsoft's performance in new fields? This is a completely genuine question. I'm not an apologist just trying to defend MS in a roundabout way.
People become obsessed with their scores and their peers' scores. They're wrapped up in internal competition and compete poorly with external players or when assessed honestly with most reality-based metrics.
I worked with MS (for a consulting company, first few years of my career) back in "the day", and I got to witness first-hand the craziness of this system. But rather than blab about my own experiences, here's a recent article that sums things up pretty well: http://www.vanityfair.com/online/daily/2012/07/microsoft-dow...
Functioning planned culture: guild culture. Master crafts(wo)men and apprentices are clearly defined, but the managerial relationship is advisory and focused on mentorship. The major downside of guild culture (the rarest of the 4) is that it relies on employee loyalty to capture the value it generates.
Pathological planned culture: rank culture. (This is the most common corporate culture.) People rise or fall based on subordinacy, rather than the quality of their ideas or their level of effort. You end up with a lot of people who don't work hard because they realize that effort doesn't matter, and a lot of bad ideas getting into implementation. Rank cultures often can't compete on an open market or address new challenges.
Functioning market culture: self-executive culture. Here you have a flat hierarchy and employees have a lot of autonomy. However, they're usually expected to take responsibility for their own advancement. This is the Valve-style open allocation culture.
Pathological market culture: tough culture. High-stakes performance reviews, low trust of employees but similarly low guidance. This is the sink-or-swim culture.
Rank culture tends to turn into tough culture as it generates underperformers, and eventually the higher-ups get upset about the whole thing and have a crackdown. However, tough culture turns back into rank culture as the people who control the performance assessment become the new rank-holders and, in exchange for loyalty, offer safety and advantage in the evaluation process. So tough and rank cultures tend to fall into a degenerate, enervating pattern of oscillation from one extreme to the other. This back-and-forth eventually exhausts the company, leading to constant reorganization and turmoil.
When a company institutes permanent tough culture (stack ranking as an inflexible pillar, rather than a temporary measure) the political corrosion associated with rank culture still occurs, but there's a selection dynamic (similar to antibiotic resistance) where the forms of rank culture that survive are the least detectable. So you end up with a culture that's politicized and gamey like a tough culture, but inefficient and extortive like a rank culture, and ultimately you have a workplace where people put in lots of hours and apparent sacrifice, but there's little getting done and there's no vision.
The team(s)/business unit responsible for XBox had a ton of independence and freedom early on and it contributed to their success. That freedom has slowly been taken away and you can see the effect in the product(s) coming out of that part of the business now.
Kinect is successful and an interesting product, but they've had a hard time producing genuinely interesting or high quality Kinect games, either first or third party. The ones that are successful tend to actually make rather minimal use of the technology, or use it in odd ways. For a couple examples, IIRC Dance Central isn't able to use the actual skeletal tracking technology, and Double Fine's Kinect games likewise take a different approach that doesn't use the skeletal tracking. Most of the games shipping labelled as 'Better with Kinect' use it as a glorified microphone, which is kind of depressing. At this point there is also a general consensus that the Kinect camera's accuracy is too low for many serious game uses and that the field of view hampers its use in many homes. In their defense, it's rumored that they have addressed this with the next revision of the hardware.
My point is more regarding things like the dashboard updates - the XBox 360 dashboard slowly turned into a glorified billboard covered in ads that has incidental features tacked onto it. They've had consistent performance problems with their system software that have never really been addressed (though thankfully they've made slight improvements here and there) - at this point people who have to interact with it on a regular basis, like journalists, are barely able to contain their hatred for the user interface.
