Slashdot Mirror
Ask Slashdot: How Does Your Company Evaluate Your Performance?
jmcbain writes "I'm a former Microsoftie, and one thing I really despised about the company is the 'stack ranking' employee evaluation system that was succinctly captured in a recent Vanity Fair article on the company. Stack ranking is basically applying a forced curve distribution on all employees at the same level, so management must place some percentage of employees into categories of overperforming, performing on average, and underperforming. Even if it's an all-star team doing great work, some folks will be marked as underperforming. Frankly, this really sucked. I know this practice gained popularity with GE in the 1980s and is being used by some (many?) Fortune 500 companies. Does your company do this? What's the best way to survive this type of system?"
The best way to survive is not to play the game.
What's the best way to survive this type of system?
It's called a union.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
"What's the best way to survive this type of system?"
Find another job where they treat you as a human being.
If you've got some serious skills, tell them to stick it and go work for a smaller company that's been around a while. Right now it's an employee's market so to speak with respect to certain technology skills (I've been off the market over a year and still get 10+ recruiters calling me a week, and I'm not all that great at all!). My thinking is that you've got more choice than they do, and that after you and hopefully everybody reading this reply, and then some, tell their HR departments that this kind of performance review bullshit is why you're leaving, things may eventually change.
If employers start seeing their very-hard-to-replace talent walk out the door because of draconian, 30+ year old management paradigms, they may be forced to change.
1. quit job
2. build start-up
3. ???
4. Profit!
5. hire jerks that gave you bad stack result
6. treat them stack performance game
7. Revenge!
The only solution is to get another job because you cant win. You can get higher up but by then all you really do is internal politics, stabbing your friends in the back and running around PR-campaigning for yourself. Work, not so much. If you really like politics, lies, distortion and stuff, get a job in politics instead of masquerading as a coder when you in reality is doing politics full time.
HTTP/1.1 400
1. create dummy identities in your team
2. make those dummies look underperforming compared to you (I know, this is the hard part)
3. next stack ranking comes, they get in the pool, so you are above average.
4. profit!!!
I believe this technique is called "stack overflow" and I bet it will work for microsoft for another 30 years at least.
---- MISSING MISCELLANEOUS DATA SEGMENT --- [sigdash] trolololol
I work in healthcare and the model often used is Brenners Novice to Expert. This looks at the development of an individual in their practice. While a great model since it allows one to compare themselves to themselves and looking for improvement, it also promotes team work. Of course this is a little difficult to apply many software firms. Another model is using a 1-5 scale, where 5 is exceptional, 1 is unsatistifactory, 3 meets criteria, 4 is exceeds criteria, and then they tally these for whatever metrics used and divide to get an average. Comparing staff to each other does not develop team work and only works in competitive environments like sales where you want people to outdo each other.
30% Troll, 50% Underrated, 10% Interesting
Score:5, Troll
At a former employer I joined a team that was under-performing. I worked hard to get things back on track and I did my absolute best. At my bonus meeting my boss told me that I had done a great job and I was the best performer on the team by far, but he had to give a certain number of people a good review, some a fair review, and one an under-performing review. He didn't do this by job performance but by length of service, and since I was a new guy he gave me the poor review so I got almost no bonus! After that I didn't work so hard....
What's the best way to survive this type of system?
Set up your own religion like L. Ron Hubbard.You could also found your own Fortune 500 corporation but that's more work. Which ever path you choose it boils down to the same truth, if you are the grand poobah you don't have to perform, only punish your underlings for not doing so.
Only to idiots, are orders laws.
-- Henning von Tresckow
Overachieving isn't guaranteed to get you a high ranking. It's a political game, much like popularity in a high school. It's not about how well you perform, it's about who you know.
Does your company do this?
Yes.
What's the best way to survive this type of system?"
Gamify. At my company, what makes things even worse is that to be considered in the top 20%, you have to show initiative and contribution *outside* of your core responsibility. This involves:
Whatever you do, absolutely never, ever get your head down for long periods and just get things done. That is the road to, at best, an "average" rating. You see, by doing your job well, you are simply doing what is expected of you. It does not matter how complex or easy your job is - no one knows or cares. All they see is someone doing their work.
This might be a difference in work-place culture, but whenever I choose a job I always only considered the fixed salary part for comparison. If I was happy with that, the job is ok. If I need some bonuses to make a decent living, it was re-negotiation time. The nice consequence of this is, that I don't care much about the rigmaroles with performance reviews to decide on the bonus. That makes me very relaxed and whatever comes in is just a nice bonus and nothing I really need. In the end by not caring, I swim along with the average, but I still can tell them to get stuffed if the idiocy becomes too rampart. And being the one to stand up and voice what everyone is thinking sometimes makes you popular or someone to be consulted beforehand.
In the companies I worked for, the more formal and stupid the system was, the easier it was to gamble. I liked best the system with self-defined yearly goals, where the road to success was in the skill to formulate impressive sounding goals where the non-performance was hard to verify. Or to be part in projects that get shut down because of reorganisation before being delivered. That never got me top rates, but before going through the hassle of digging through the bones for some real data average success and bonus (or slightly above average, if I bickered too much about my valuable contributions) was assumed independent of the actual performance.
For me that gives the best results for a minimum of exposure to the whole idiocy.
This ensure that the company only keeps burnt our overachiever and political sharks.
With a little bit of luck it'll drive them into the ground.
And anyway anybody working for microsoft deserves "advanced corporate management techniques" being applied to him or her.
Don't put up with that shit. Vote with your feet and quit your job. Stop being a bitch.
Or, depending on your region work opportunities and your physical appearance, start being one.
I've been on a company that follows this aproach
They had a fixed percentage to put into above the average, result: ... ...
- some years a few good people had an average evaluation -> good people get frustrated -> good people leave
- some years a few average people had a good evavluation -> good people that last year got an average evaluation get frustrated
They also had a fixed percentage to put into bad evaluation, result: ... ...
- some years a few average people had a bad evaluation -> average people get frustrated -> average people leave
- some years a few bad people had an average evavluation -> average people that last year had a bad evaluation get frustrated
From year to year there seems to be a group of good people, lets say 30%, that try to get into the 20% openings for "good evaluation" the result is 10% will always get frustrated and consider the system unfair because they consider themselfs above average.
Everyear, a lot of people (perhaps 40%) are not incentivated to fight for good performance because they are not good enough or not willing to sacrifice personal time/life to achieve that mark, nevertheless they are good enough to have an average evaluation, they just go with the flow ... ...
For these people the system has no impact whatsoever
All in all, the system seems to have some advantages but I'm not sure the advantes are greater than the disavantages.
How to survive? ...
What are you aiming for?
To be on the average evaluation, you usually don't have to do much, after all you must be better than the bottom 10% or 15%
To be on the above average evaluation if you are on a star project on the company and if you are good you have a good change to get a good evaluation, if you are on a marginal project, don't even try it you will have to work 2x has hard as someone on a star project to get a good evaluation.
And offcourse there are politics, some people will get a better evalution not based on performance but on social connections ...
Between the people that don't care if they get fired, the people that won't fight to get to the next evaluation level and the people that would fight to have a good performance even withough the system in place, this system is only having impact on 25% of the people.
I think we're still a Fortune 10 company... we manufacture consumer products globally, and have a global performance evaluation (PE) process. I will be as generic as possible in the terminology. Oh, I'm a manager who conducts PE's, and also a volunteer on the personnel development forum (PDF) for non-management personnel.
For PE's, we have a top-tier level that's limited to 15% of the eligible pool. In my department so far this year, we've not nominated enough people to meet that 15% (we're in a new region, and all of the local employees are new). Then there's 70% to 85% of people that are achievers. This bracket is slightly open because there's an allowance of 15% of under-achievers and non-performers. The key is, we're *not* forced to bracket anyone into the lower tiers. And like I said for the top tier, we're not forced to bracket people into that tier, either.
Our system makes sense. Not everyone can be a super-star; even when everyone is a super-star, there's always a small percentage that have a little bit of an edge. And because we're not forced to rank anyone as under-achievers, we recognize that even the weakest link might be carrying his or her weight -- and carrying weight (do your job) is all we ask!
To prevent abuse, all of the top 15% and the lower 15% (if any) all go to the PDF committee that I mentioned I'm in. There, my fellow managers and I review the proposals for the highest-achiever rankings, and we all have to agree. Basically, you can't screw your way to the top, or other methods of brown-nosing.
And as a low-level manager (organizationally-speaking), I'm subject to the same process at my pay grade. And I'm fairly happy with it.
--Jim (me)
This pretty much guarantees a rat race.
Everyone tries to get ahead of each other. This can work, if the metrics are totally objective, but they rarely are. Even if they were, it's a high stress way of doing things. Only a certain personality type thrives under constant stress. These people will usually leave and set up their own business
Also, MS hires good people. If you are competing against other good people (not useless dolts), then it's hard to win on ability alone. It's far more effective to do a reasonable job, and suck up to your boss / make your boss look good / advertise your "achievements" to your boss's peers, etc.
Eventually, the people who are good at the game get promoted, and forget that the game is actually a bad thing. They start consciously rewarding people for playing the game (not getting fooled by it, but actually expecting their workers to game the system), and madness prevails.
On my team, we use a sort of 360 degree review process. The people I manage meet with me and my boss on a quarterly basis, and we use the time basically to check in on any issues we've identified, check on any goals set in the last quarterly review, talk about training / certification progress, listen to any concerns they bring up with people / processes / environment, etc. At the end of the review, I leave the room and the employee gets to talk to my boss about me, without me in the room. Then I come back in, my boss leaves, and we talk about my boss without him in the room. My boss and I aggregate and anonymize the top 2-3 things that people mention about us, and that becomes a part of our reviews.
Within: Is to remember that the ratings are subjective. Make friends. Particularly make nice with the boss. Then perform competently so they have no reason to downrank you, while having reasons to rank you above the other competent people.
Without: Is to leave for a place that uses a sane management system. There are plenty. Some of them are eating Microsoft's lunch right now. People who are actually competent software engineers are in extreme demand right now, there's no shortage of jobs for that skillset. A recruiter can get you a list of a few hundred positions for you to choose from on a moment's notice.
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
... is be ranked the best.
FTFY.
Every end has half a stick.
This doesn't work though.
I can give you a real world example. My partner was a retail manager for a well known denim retailer, and was consistently the top in the country each year in terms of year on year growth no matter which store she was assigned to. The problem is only one manager was allowed to be graded 1 on a scale of 1 to 5 or whatever it was. She obviously deserved it as she was the one consistently performing the best, but this had issues.
If it was given to her year on year, the other managers felt they had no chance and just had no motivation to excel themselves because they were only going to get a 2 anyway - they could try really hard and almost do as well as her, but not quite, so why try when they'd get a 2 regardless? The worst part being these grades were linked to your annual rise, so no matter how hard you tried you'd only get a smaller rise against her.
So the management figured well hey, we need to motivate the other folks, so we'll give it to them for how hard they tried, rather than actual results, and so then it's my girlfriend who despite her stellar performance instead suffers and gets a lower payrise. Then she has no motivation to really continue to outshine, because she's not gaining anything for it, in fact, someone that performed worse than her is getting a better reward than she did.
Of course, the solution may then be to increase the number of people who get a 1, but then at that 1/2 boundary you have the EXACT same problem. Those who try real hard can never quite catch the top performers so are not motivated to do so. Really, the only solution is to instead rank people based on a sensible balance of effort and performance without any cap on how many can be deemed to be top performers. If you extend it to say 5 out of 30 people can have the top grade, then what happens when your survival of the fittest type management system gets the 6 best managers in the country in? well, the 6th one will fuck off elsewhere because they'll be getting shit on relative to the others. It'd be far better if all 6, or all 7, 8, 9, or 10 could get equal reward so that you retain the 10 best managers in the country, rather than stick yourself permanently with only the 5 best, letting the other 5 fuck off to your competitors.
The problem with your theory is that yeah, it works great for the top person, but what the fuck is the use in a system that means that 29 of your 30 staff just have no reason to be motivated? That's a complete failure of management.
I've personally not had a problem being in the top percentile myself, but I'd absolutely fucking hate to work under this system because it'd mean everyone around me was unmotivated meaning I'd be carrying the team - I want my coworkers to be motivated, I want them to do well, to be praised, to be given reason to care about their job, because that makes my life easier regardless of whether I'm a top performer, or a bottom performer.
At the end of the day these braindead systems exist because of inept managers who are either hiring the completely wrong people, or don't have the balls to tell someone truly inept, lazy, or incompetent that they're fired. These managers either can't actually figure out how well their staff are performing, or how competent they are, or they can, but just don't have any spine to do what's required to deal with them, and so they give them this absolutely failure of a crutch to try and automate the process for them but it merely serves to destroy motivation of those who can perform by giving them reason not to.
If you hire good managers you don't need this kind of crutch, the good managers will know who deserves what praise and reward, and who needs to be fired.
Most competence systems I've seen has a 1-5 performance rating where 3 is performing okay, 4 well and 5 exceptionally. On occasion there's a 2 for underperforming and very rarely an 1 which is basically fail but it's rare because you shouldn't get promoted to that level if you aren't already performing like one. That's usually reserved for total mishires or people who've had some kind of personal breakdown. Saying that X% of your workers are underperforming is saying that your hiring process fails X% of the time - that figure should be close to zero.
Of course before that there's usually a set of skills that your employment level should have, so the demands on a "Senior Developer" is different from a "Junior Developer". Usually these are set up in a competence matrix, so when they're looking at possible promotions they can say yes, you're coding at a Senior Developer coding level but you lack skill X which is required to be a Senior Developer. Skills development is related but actually quite distinct from your work performance, you can have done your job excellently but done very little to improve your skill set.
Sane companies also look at professional development, if you're a first year Senior Developer whose performance has improved but still is below average you're probably a better choice than the 5th year Senior Developer whose performance has declined and is now equal to yours, those two are connected. It was probably a better idea to promote him to an okay performing Senior Developer than for him to be an overachieving Junior Developer. That's another reason 5s are so rare, if you are that good you should be in a position with higher demands.
That said, when it comes down to it managers can pretty much manage to tweak the rating however they want. That said, even the worst of managers want to look good to their team/departments bosses and customers. If they know you're critical for them to deliver on time and in good quality, you'll survive most of the office politics. But without trying to kiss too much ass, make sure your boss knows what you're doing for him. Don't expect him to find out on his own.
Live today, because you never know what tomorrow brings
Network.
Find the smarter people who are better-connected than you, ask if they know why your idea is dumb. Get them thinking about it. Let it percolate.
-- IANAL, this isn't legal advice, and definitely isn't legal advice for you. Also, Squee!
Most of the problems raise when big numbers are translated to small teams. Probably, a 10% of a big corporation staff deserves to be fired, but that's not appliable to 10% of every team. Maybe several teams deserve to be fired alltogether (boss included) and some teams deserve an extra-bonus (ok, boss included), but the big numbers should achieve those global percentages.
Bad bosses apply the corporate percentages top down without changes because it is easier to manage and they can say "I'm not mean, it's the rule" but he is part of the problem. The spreading of the percentages should be distributed through the organization weightened according to the contribution of the teams and sub-teams, so there could be an uneven punishing policy, which is counterintuitively far much more fair.
First-line managers who have to deliver against HR policies like this have my every sympathy. I was made a manager in a certain very large IT company. I managed a team of mixed fixed-term contractors, contractors and permanent staff. My manager came to me at the start of the new year to tell me that during the upcoming staff performance review, I had to make 15% of my permanent staff a 1 performer, 75% a 2 performer, and 10% a three performer. When I complained that I didn't have enough permanent staff of a low caliber (c'mon now, I was doing the hiring!), I was quite neatly told that if I couldn't make up the numbers from my workforce, then it would be OK as from his level he would meet his overall target for 3 performers by making ME one.
Actually, that's what ended up happening, not that any of my workforce found out about that 'deal'. I lasted a further four years of management in increasingly Kafka-esque circumstances until I decided that I should stop trying to rise up the ranks of management, give up and go back to being a techie. I've never regretted the decision, and I can sleep at night.
I have no idea how my current employer does performance review. I haven't had to deal with performance reviews in over 15 years. This is one of the benefits of working as a consultant on a contract, and one of the things I don't miss about working as an employee.
I personally find consulting to be a more civilized, sane way to earn a living. My total compensation gets negotiated up front, for some prescribed period of time. Then, when the time is up, we just negotiate again(1). Simple. No fuss, no mess. You know how much you're making, and you don't feel shortchanged when the bean counters decide to cut down on some fringe benefit.
I guess that periodic contract extensions would count as a periodic performance review, of some sort. But there's no bureaucracy involved, and I don't need to dance like a pony, in front of someone. It's purely a business transaction, and nothing more.
The oft heard suggestion of unionizing is a joke. It's never going to happen. If you want to unionize, sure, but good luck to you. On the other hand, if you want to become a consultant, that can happen today. Your choice.
(1) Yes, I've went through an occasion of an 800lb corporate gorilla deciding, by fiat, to cut all their consultants' rates, for budgetary reasons, assuming that everyone is going to accept it and that they have no choice in the matter. As my then-managers discovered, that assumption was wrong. One of the other benefits of consulting, you see, is far fewer questions of what happened at your last job. Naturally, contracts come to an end all the time, and one's services are no longer required. Nothing wrong with that. Perfectly understandable, and expected.
Only it doesn't work, because this type of system ensures that only the top percentile remain in the team in the long run, meaning that last year's top performances become this year's average target. In short, if you're an overachiever, you'll raise the bar for everyone else, including yourself. It's a self-defeating system.
On top of that then there's the issue of actually measuring who the top performers actually are. Ok, with some jobs that might be fairly straight forward. However most of us here are techies, how do you measure the performance of a software engineer or quality assurance. (I ask because I know some of the people in the company get rated highly because their managers talk them up and the company gets this fantasy about how great they are. Should I mention I have a QA person now that basically had me waste months of company time "fixing" stuff that wasn't actually broken? But since this ones a top performer there's no way I was allowed to not fix it.)
Did you know 80 to 90% of the moderators on slashdot wouldn't recognize a troll even if one dragged them under a bridge.
Dont forget to back stab those around you. You are competing against them, not working with them for the greater good.
That might explain some of the M$ bugs we have seen.
im not overly popular with typical HR people, cause i ask lots of questions during the initial interview, such as: how do you handle performance reviews? how is management reviewed (top down, 360, etc)? it weeds out bad places to work real quick. plus, its really fun to see the look on an interviewers face when theyre put on the spot.
I got introduced in this system twelve years ago. In the nineties I had never worked for a company which did such evaluation.
I have no qualms about evaluations per se, but when I heard how this worked, my immediate reaction was a real WTF moment.
I have in the course of school and my career been introduced into statistics several times, and I know the Gauss curve. So my first reaction really was, wtf. you do not go measuring and plotting your data, and then expand your bell curve. No, if you want to know if there are outliers then you do this match against your mean and your standard deviation. That way you can see the underperformers, but also the people who are really, really good (or one should investigate the matter).
However, the biggest wtf is really that I am working in a company with many engineers (master level engineers). I expect these people to understand these issues in probability/statistics and made a statement against this misuse of mathematics a long time ago, which is absolutely not the case.
I believe my company has the same system, only the complete gutting of bonuses years ago and the fact that a stellar rating gets you around 0.3% more on the pathetic annual salary increase means that no one cares.
My last performance review contained two directly contradictory statements from my manager, in what I suspect was an uncorrected cut-and-paste from the previous year. I didn't bring it up, because either way I was getting the same shitty raise. That's motivation for you.
My primary workplace uses the 1-5 scale, but doing a perfect job of everything you're asked to do gets you a "3 - meets expectations." So the only way to get higher than a 3 is to think of things to do, and do a perfect job of them, before it even occurs to anyone to ask you. I averaged a 4-of-5 across several categories on last year's annual performance evaluation, which got me a year-end bonus equal to about 2 weeks' pay (which we tend to get most years) and a raise of... nothing.
Village idiot in some extremely smart villages.
"Your GF should have been promoted to a trainer of other managers, given a raise and been told that she now only meets expectations and is competing with other top retail management trainers. Now she's responsible for creating more copies of herself."
Which is exactly what she ended up doing. But how does that help the problem? someone simply fills her place and you still have the exact same problem again with the person who filled the space left by her exit.
"They also need to break that group up. Make it a regional challenge so there are no more than 10 peers in a group at least. 5 is ideal. Then set the bar for top performance at or just above the best in the group the prior year."
So what happens when you have a region where there happens to be 3 extremely talented managers? You've still got the exact same problem.
It's a broken system, you can't fix it by changing the parameters, the fundamental problems all still exist. The only solution is to use a more rational and forward thinking method of grading people that does away with these fundamental flaws altogether. What if someone has had the odds stacked against them because there's been a flood in their area but they still managed to break even? do you punish them for not achieving growth? The point is this - everyone's circumstances are going to be unique, so lumping them all into the same system is a recipe for failure - someone is always going to get treated unfairly and that has completely the opposite of the desired effect, it demotivates them. You can't escape the fact you need competent managers that can judge people based on their, and their store's unique circumstances.
At the end of the day these braindead systems exist because of inept managers who are either hiring the completely wrong people, or don't have the balls to tell someone truly inept, lazy, or incompetent that they're fired. These managers either can't actually figure out how well their staff are performing, or how competent they are, or they can, but just don't have any spine to do what's required to deal with them, and so they give them this absolutely failure of a crutch to try and automate the process for them but it merely serves to destroy motivation of those who can perform by giving them reason not to.
I think the term "manager" has been severely overloaded in the modern corporate world. I think there are at least three different types of management:
1) Budget management - bean counting and other forms of paper-pushing
2) People management - acting as a buffer between your team and the rest of the company, as well as dealing with issues internal to the team like personality conflicts, motivation, career path, etc
3) Product management - mostly technical and creative work all revolving around the life-cycle of a product from creation to end-of-life.
I think that a lot of the dysfunction in modern corporations comes from the intermingling of these distinctly different jobs under the common rubric of "management." You end up with people applying bean-counting processes to people-handling situations and the result is nearly guaranteed failure.
When information is power, privacy is freedom.
When it comes down to it EVERYONE has their own business. When you are traditionally employed, your business has one customer, and if you lose that customer by quitting or getting fired, you're out of business. Start your own business and remember each customer is an income stream. Multiple income streams mean more money and more security, and also give you the ability to fire customers you don't want to do business with.
This doesn't mean it's easy or even possible for everyone. My business was much harder to start than I ever thought it would be, but the challenges have been worthwhile both in income and in getting out of corporate BS like the stacked ranking game.
Middle managers who have no skills beyond playing office politics and self-promotion are pretty much stuck in the corporate rat race, but people with real skills that translate to marketable goods or services can make it on their own if they can learn how to build business structures and processes to run their business and a marketing plan to get customers.
The Slashdotter who said the best way to win is not to play the game was right. This post suggests one way HOW get out of the game.
I've worked for big companies (IBM) that had twice-a-year reviews that involved goal setting, evaluation of whether goals were achieved, and rating yourself on a bunch of silly categories. Bonuses were tied to your score. I've worked for small companies that tried to do the same, and I've worked for even smaller companies where there was no review system (or performance bonuses) whatsoever. I vastly prefer the latter. If I'm doing a crappy job and am in danger of losing my job then tell me. If I'm doing an awesome job and you're especially pleased with my performance then tell me. If I'm meeting expectations but not doing anything awesome then don't waste my time (and create awkwardness between manager and employee that needn't exist) by making me go through performance reviews.
> To survive in an environment with a forced bell curve, let me paraphrase Shakespeare: "Kill all MBAs."
The people who do nonsense like this are often not MBAs. MBAs should be telling them why this is a bad idea. In my experience, MBAs are taught that this is a bad idea.
> if the company did its job correctly in the first place (hiring great people and not hiring bozos)
Unfortunately, this is hard. Sometimes you make a mistake and hire a bozo. Sometimes you hire a star, and he becomes a bozo. Personally, I like the motto "Hire slow, fire fast." which I think goes to your doing "its job correctly in the first place." Take the time to hire the right people. When you end up with a bad apple, don't take a year to get rid of them, especially if you have a 90 day probationary period.
Actually, that might be one thing that leads to systems like this. It's hard for a big company to fire just one person. They tie themselves in HR knots. But if it's standard operating procedure to fire X% every year, well that's easy! I'm not saying I approve, merely that I may see why this particular pathology develops.
My company used to do this and yes, it does suck. Not only because it does force unnatural rankings depending on the mix of people, but because the good old boys and people who have connections don't get weeded out as part of the process. I remember one of my good employees getting 'targeted' to land in the bottom of the rankings and having to haul the rank meeting manager and the HR person into a different room and asking if they wanted to continue tarring and feathering the good employee, or should I go back in and bring up the couple of 25 year plus employees who did nothing more than recirculate the air in a cubicle. Turns out they didn't.
How do you defeat this? Pretty much perception, perception, perception. To succeed in one of these things, the managers in the room folding, spindling and mutilating your annual contributions should all know who your employees are, approximately what they do, and have a favorable impression of them. This is a year long marketing effort to get recognition for your people, name them in staff meetings and in written status reports when they do something good. Death is some manager in the meeting that one of your employees did something to during the year, but they decided to wait until the review process to bring it up.
The difference between the guys at the top and the guys at the bottom are the ones at the top got talked about and everyone in the room said "Yep, good guy" while the ones at the bottom were people nobody knew, someone had a bad experience with them, or nobody understood their accomplishments.
So the marching orders are a) make sure you know what you're working on and that what you're working on has measurable value and is important to the business. If you cant identify the value and importance, simply stop doing it. Make sure everyone knows what you're doing. Make sure every time you interact with a manager that its a positive outcome or bring it up with you so you can repair the situation in advance of the review session. Market the heck out of your people and put them in front of as much management as possible. I used to send employees in my stead to meetings or have them make major presentations where most managers want to do it themselves.
Done properly, this could be a good tool. Not done properly (and it usually isnt done properly) its a stress inducing sales job and whoever has the best skills at presenting employees and hardballing the HR people will get the results.
There are also a number of other little things to pay attention to. I found out that each of these sessions has a hunk of money and stock options to give out to the group, but that they rarely allocate all of it and if it isn't allocated, that falls back into the general pool. So I found out that if I approached the HR person and asked if there was any residual we could divvy up among the top 2 or 3 people, they'd often do it.
When I joined LM I was already cynical about these performance evaluations, so I tended to ignore it, but a friend of mine who joined around the same time took them very seriously at first. He worked extra hard and documented all of his accomplishments in detail. He made a very strong case for his excellent performance. He ended up with the same 3 everyone else got. His boss was honest/dumb enough to tell him straight out that he gives everyone a 3 because LM needs to keep the rates down to remain both competitive and profitable. My friend then adopted my attitude. I've moved on to better things and he is still with LM comfortably coasting. LM lost 2 high performing professionals as result. One left and the other quit trying.
Very often, people confuse simple with simplistic. The nuance is lost on most. - Clement Mok
The first thing is that, as a manager of a small team, you do NOT have to meet a curve. That's only required at high levels with hundreds or thousands of employees in the pool. You DO have to rank your people in order and argue for them at a meeting with your peers. If you have a team of 6 or 8 people, I'll be very surprised if you don't know who your best person is--and who the worst one is. As a general rule, you ought to be able to rank your whole team in order from best to worst, with perhaps a few ties. (Generally, though, I didn't end up with ties.)
So together with your peers, you now try to slot 50 or so people into three rankings: 4.0 for the best 25%, 3.5 for the bulk of the people and 3.0 for the bottom 20%. (There is special handling for superstars at 4.5 and total losers at 2.5, but that's a post-process with no quotas.) The argument always revolves around strong 3.5 people who "ought" to be 4.0 and weak 3.5 people who "don't deserve" to be 3.0. Not a surprise; every manager overrates his/her own people. The pressure to meet a quota forces people to have hard arguments about how valuable each person's work really was. It can even help a manager see the importance of putting people on the highest-value tasks. At the end of it, there are typically two or three borderline individuals, but everyone else pretty much has the rating they actually earned. The General Manager takes the result up to the stack ranking at the next level, armed with appropriate arguments for the borderline folks.
One time, I worked on a project with high-visibility and lots of pressure. At review time, we told management we wanted to give about 50% 4.0 (instead of the usual 25%) and only one or two 3.0 reviews (out of a team of ~100). They pushed that up, and it was granted. We did exceptional work, so they let us blow out the curve. But it only happened once in 14 years.
What are the alternatives? Have a Union that gives everyone the same rewards regardless of the work he/she did? Doesn't seem like a winner to me.
So to answer the OP's question, how do you succeed in such a system, the answer is: work hard, do good work, help others who get stuck, and BE SEEN DOING IT. When your manager says "Jane is my best worker," you want all his/her peers to nod and say "yeah, Jane is great! She helps us out all the time!" When your manager says "Jack deserves a better rating," you don't want his/her peers to say "that lazy bum? He couldn't find his ass with both hands!" But most important of all is for your manager to actually see you as someone who gets stuff done. Whatever anyone tries to claim, most teams only have a few such people on them. They rarely go unrewarded.
--Greg
Wow, this really strikes a nerve. I worked as a financial systems developer for a large construction firm for close to a decade. They had a fantastic culture, it was a really great place to work; until our CIO was replaced by someone from GE who promptly implemented stack ranking. We literally wasted ~20 hours per person over a couple weeks working on (and refining) our "internal resumes". I would consider my ten person team to be very high performing, and at the end of the process, I was basically told that I was mediocre, and needed to step it up or risk serious repercussions. A few weeks later, and after demoralizing just about everyone in IT, she announced a "voluntary separation" program. Coincidentally, that same day, I pulled the trigger on another position (at a large software company) and I turned in my separation paperwork 15 minutes after I received the forms. I was denied for the program because I was an "exceptional performer". I asked how I could be both mediocre and exceptional at the same time, and never got a good answer. Of course I left anyway, and never looked back. The new CIO didn't make it a year.