6 ms·

The Push Against Performance Reviews

60 points by dankohn1 11y ago

analog31 11y ago
“Given what I know of this person’s performance, and if it were my money, I would award this person the highest possible compensation increase and bonus;” ... “This person is ready for promotion today.” I was a manager through a few performance review cycles before I went back to engineering. This is the first instance I've ever seen where the performance review gathered information that was actually germane to the outcome. Managers have incentives to inflate appraisals... Absolutely. My salary budget was incremented in the same way as our salaries are, by getting the biggest raises possible. I was competing with other managers for a piece of a fixed pot of money for these raises, and every manager knew what was at stake.
HelloMcFly 11y ago
I try to make the case like this in my organization (and have in my previous one as well): we can easily quantify the enormous amount of money - particularly but far from exclusively the financial value of time - we spend on our performance review process. But what we don't really know, and what very few organizations really know (because "performance" is usually a very troublesome criterion) is what value that process is delivering back to the organization. One of the problems is that most organizations want their process to "Do It All." They want it to be development for the employee while also being the primary determinant of administrative decisions. Further, they want it to support all administrative decisions: who's great at their current job, who is most capable of succeeding in their next job (i.e., who should we promote?), how should we distribute our pay raises, what do we do with this year's STI and LTI, etc. It's broken, and expensive. Maybe it can't truly be fixed, but at least we can save some money, time, and frustration. That doesn't mean nobody gets feedback, or that we don't set and track goals. But it doesn't have to be the way we make it. If we put half as much effort into making managers better at leading people I have no doubt the impact would be much more clear.
- eitally 11y ago
  Imnsho, at most companies the process only serves as 1) an atta boy for the high performers, to justify raises & promotions, and 2) CYA documentation to manage out the lowest performers. These extremes cover about 10% of any org, max, and the rest are all completely unserved by the process.
LoSboccacc 11y ago
people optimizing for the metric instead than getting better at working? color me surprised. this has been know for a while by now: if you try to put down hard metrics for creative jobs they just backfire. also managers using reviews to favor their protege and as favor exchange across other managers to partition their resources shouldn't come as a surprise to anyone having worked with humans.
perlgeek 11y ago
The problem is, those existing systems were there for a reason. And I'd guess the reason usually was that the previous, less formal system also had their weaknesses. Most notably, they are very probably very subjective. IMHO performance ratings and deciding who gets a promotion simply are very hard problems, and no good solutions. So people are always dissatisfied with the current approach, and so add rules and layer to fix it. Eventually, the system becomes unworkable and bloated, and are stripped down to the basics again. Which we currently observe. But I wouldn't be surprised at all if the cycle continues, and more rules and formal processes will be added.
- yuhong 11y ago
  I do think trying to quantify performance into a number is flawed for anything other than factory work and the like.
  - perlgeek 11y ago
    Yes. And the dilemma is that there has to be one number (the one on the pay check), and it'd better correlate with performance in some way.
    - yuhong 11y ago
      I think a better approach is to give the same salary to an entire team.
      - eitally 11y ago
        Why do you think this is a good idea?
      - solomatov 11y ago
        The more creative the work, the higher the difference in performance. It seems to, it's unfair that people who do substantially more, are paid the same amount of money as underperformers.
kfk 11y ago
I think it's good to sit and talk about performance and work every now and then, I don't think it's good to write things down and put them into a system for years. Also, I think it's horrible to have things like OLR where "leaders" sit and talk about you and how well you do (from below to above expectations). The thing is, most managers are bad, especially the middle ones, and their opinions about their reports tend to be bad and flawed. Managers get overly excited about extroverted people and the introverted struggle because they don't show "leadership skills". However, in many fields (i.e. mine in finance), you need a high % of introverted people because they are the ones that usually perform the work, while extroverted spend time socializing and "networking". This whole situation creates a continuous incentive to game the system - instead of doing your job, you spend time working out ways to game the reviews (there are many, the classic is when you are indispensable for some mission critical tasks). The result of this is that you will get politicians at the top, those that game better the system, instead than results. It's not easy, but definitely performance reviews and OLRs are not the answer. I think 1:1's are a good idea, because they happen more frequently. As for feedback from others than your boss, I think it would make much more sense if that was on the whole team/department, rather than on single individuals (like in OLRs).
- zekevermillion 11y ago
  The extrovert vs. introvert distinction explains a lot. I'm an introvert, and have always hated reviews, even when I know it's going to be a good review. The best review in my opinion is when I get no response at all -- that's when I know I've done my job, and it just worked. Extroverts seem to enjoy getting "feedback" even when it doesn't matter -- e.g., when there's no raise in it, and there's also no chance of being fired. Unfortunately in most companies, compensation (even that which is nominally based on performance) is not based on performance, but rather on the market. An accurate comp-linked review would go something like: "we decided to give you a 1% raise this year b/c we think that is the minimum necessary to keep you in this job; also we used up our budget to hire this other person, and she needs 10% more than you're making to convince her to leave her other job."
  - ild 11y ago
Intermernet 11y ago
Performance reviews are a form-letter attempt at managerial automation. They seem to suggest that a manager should only be monitoring the performance of their team at certain points in the year. A good manager (a rare breed) knows to a fairly certain degree how their team is performing at any time based on the output from the team. If they don't know this, how are they meant to manage? I flat out refused to do performance reviews at a previous position. If the company has an issue with my performance, then HR can talk to me directly. If the company wanted to fire me for not filling out their performance review, then I didn't really want to work for that company. They never mentioned it, and I left the company amicably.
dbg31415 11y ago
Without a formal time for review, how will they build a formal time for pay evaluation and bonuses? I can't help but think this is like the latest cell phone craze... yes, your bill drops down slightly, but you are going to have to pay for your next phone and when you do your bill will actually increase. Suckers. Stack Ranking was and is fucking horrible. And getting rid of that for honest 360 reviews should be the goal. Reviews from your peers with project completion, reviews from your manager with year completion... those are both important and can help you grow. Not to mention get a raise.
VLM 11y ago
“This person is ready for promotion today.” I would wager the percentage of employees working at a firm that still participates in promotion culture is low and dropping. In other words "ready for a promotion" for most people in 2015 means they're going to quit soon and get a higher level position at another firm. I'm sure there still exist individual companies and maybe unusual and rare fields where promotion culture still exists, but it seems very rare and unusual. Meanwhile in most areas outside SV/NYC computer programming during bubbles, there is a large excess of qualified workers vs positions. For example in K12 education the state-U system graduates twice as many teachers as there are open positions for all new and past grads to apply for. It doesn't matter if they are in the top half that got hired because they are the best, or the most politically untouchable, or the most attractive, a year after hiring at review time they're still going to be that top half. So there's nobody to give a hard time to in a review, because that lower half or lower 3/4 or whatever in any professional field are already working at Starbucks or as a waiter or bartender. The world has "needs improvement" kindergarten teachers, but they don't work at your neighborhood school, they never got a job in teaching and they serve you beer at the bar, for example. The economy is in permanent contraction and the general population and higher-ed sector are in permanent expansion for many years now, outside bubble segments of the economy, so hiring one of the bad workers strongly reflects on a managers interview skills, making it even harder to admit anyone is below "excellent". Meanwhile increased labor mobility for the few that are employed means a review a couple years back, is a couple employers ago, so it is inaccessible and irrelevant, so its not like the wasted effort today will pay off in ten years. Maybe a review in 1950 would come in handy at the same employer in 1960 to select a new director, but the firm has no loyalty to the employee and therefore the employee has no loyalty to the firm, for at least three human generations now, so its been culturally bred out of us. So the folks at the top of a team will interpret an attaboy as "start applying for higher level jobs elsewhere" and economic realities mean you don't have truly low performance employees to crack the whip upon, making the whole dog and pony show a waste of time for all involved. This leads to an attitude of just pencil whip the process to get the meaningless metric of "accomplished reviews" up to 100% and be done with it. Once the folks who've been pencil whipping it for their whole careers make it up high enough, its finally politically possible to discuss getting rid of the waste of time. So here we finally are.
- 11y ago
solomatov 11y ago
It's very easy to criticize performance reviews. Does anybody know about more sensible alternative to this system?
- greggman 11y ago
  I'm curious too. One company I worked at had me privately rank coworkers. No Idea if it worked. It seemed like it might identify top and bottom performers?
lectrick 11y ago
In a creative job (such as programming) it's extraordinarily difficult to evaluate how valuable someone is, but I would suggest some sort of human valuation ("you know it when you see it"), a LIGHT metric used only as a data point (goals and whatnot, any measurable impact on bottom line, or website performance, or some other numbers), and some means of simply measuring how much someone cares about their job, company and career. I think I can tell how much the people I work with care about their job, company and career, I think I could trust most of them to know that about me, and I think that's a good place to start. Metrics-wise perhaps a "spread spectrum" approach that covers so many things that most signs of being "productive" would end up at least reflecting in the numbers, yet would prevent "gaming" of the numbers since so many things are measured. Used only to inform and not base hard decisions on. This would catch both people who care about their job but for whatever reason aren't producing anything measurable, as well as people who have produced but perhaps could produce even more if they were more engaged.
solipsism 11y ago
The yearly performance evaluation is a joke. Any company which respects its employees as humans and which takes seriously its responsibility to its shareholders will promote the ongoing evaluation of employees based on close 1-on-1 relationships between manager and direct report. And the quality of the performance evaluation at any company that does this is only as good as the quality of those managers. If all your managers are people who are out to make themselves look good, or out to keep their own jobs safe, no system could possibly get the most out of your employees. If your managers are motivated to care about the company and to make its goals their own, and if you provide the proper training and evaluation of their management to ensure they know how to manage the right way, you have a chance.
markbnj 11y ago
I'm happy to see this become a topic of conversation. The formal performance review is emblematic of the key difference between small and large organizations. A small organization will usually rely on ad hoc information gathered from the process of daily interaction and observed results. Opinions form, are solidified, and ultimately business decisions are made as informed by these ad hoc impressions and the resulting attitudes. For the best small organizations these approaches work very well. Managers have enough personal oversight of the company's operations to feel like they have a good handle on what's happening. Once an organization grows to a certain size management begins to lose that level of personal oversight. No one person or even a small group of persons can comprehend the scope of the company's operations. More levels of hierarchy have been introduced; there are more managers responsible for more people. Now it is probably possible that those same ad hoc approaches that worked in the small organization could continue to work at some unit level of the larger organization. Managers could maintain personal oversight of their unit, the observed results could percolate up the hierarchy, decisions could be made based on those results. Unfortunately, the insecurities of higher level managers will not usually allow this to happen. Lacking that personal oversight at the operational level they attempt to compensate for the risk of a bad outcome by developing rules which ensure a good outcome. This happens so reliably as organizations grow that the emergence of rules is nearly synonymous with the concept of growth. However, and I think this is the key point, it seems to me that developing rules to ensure that an organization is managed to a good outcome is a little like trying to develop rules that ensure a baseball team will win a majority of its games in a given season. Rules of this sort actually cast in concrete a specific view of and response to a set of conditions that exist or existed up to the time when the rules were created. They stifle the ability of the organization to react to changes in conditions that may not have been anticipated. And for whatever reason, probably due to essential human nature, they seem nearly inevitable once an organization reaches a certain size. It would seem to make sense to act more like the baseball team: acquire the absolute best people, give them the absolute best training and resources, make the mission absolutely clear, and then reward or punish based on results. But to take such an approach requires waiting for those results to come in before you can know whether you were successful. Which is, interestingly, how military organizations have to handle field commands. When battle looms the hierarchy (and not inconsequentially the troops) has to put its faith in the people in command on the ground, whom they have put in place, trained, and supported. The tides of battle change too swiftly, and the outcome of poor or slow decision making is so brutal, that a top-heavy approach to governance simply doesn't work. In some ways that is the essential story of the way the war in Vietnam was fought, for example, as contrasted with previous conflicts.

60 points by dankohn1 11y ago

“Given what I know of this person’s performance, and if it were my money, I would award this person the highest possible compensation increase and bonus;” ... “This person is ready for promotion today.” I was a manager through a few performance review cycles before I went back to engineering. This is the first instance I've ever seen where the performance review gathered information that was actually germane to the outcome. Managers have incentives to inflate appraisals... Absolutely. My salary budget was incremented in the same way as our salaries are, by getting the biggest raises possible. I was competing with other managers for a piece of a fixed pot of money for these raises, and every manager knew what was at stake.

HelloMcFly 11y ago

I try to make the case like this in my organization (and have in my previous one as well): we can easily quantify the enormous amount of money - particularly but far from exclusively the financial value of time - we spend on our performance review process. But what we don't really know, and what very few organizations really know (because "performance" is usually a very troublesome criterion) is what value that process is delivering back to the organization. One of the problems is that most organizations want their process to "Do It All." They want it to be development for the employee while also being the primary determinant of administrative decisions. Further, they want it to support all administrative decisions: who's great at their current job, who is most capable of succeeding in their next job (i.e., who should we promote?), how should we distribute our pay raises, what do we do with this year's STI and LTI, etc. It's broken, and expensive. Maybe it can't truly be fixed, but at least we can save some money, time, and frustration. That doesn't mean nobody gets feedback, or that we don't set and track goals. But it doesn't have to be the way we make it. If we put half as much effort into making managers better at leading people I have no doubt the impact would be much more clear.

eitally 11y ago
Imnsho, at most companies the process only serves as 1) an atta boy for the high performers, to justify raises & promotions, and 2) CYA documentation to manage out the lowest performers. These extremes cover about 10% of any org, max, and the rest are all completely unserved by the process.

LoSboccacc 11y ago

people optimizing for the metric instead than getting better at working? color me surprised. this has been know for a while by now: if you try to put down hard metrics for creative jobs they just backfire. also managers using reviews to favor their protege and as favor exchange across other managers to partition their resources shouldn't come as a surprise to anyone having worked with humans.

perlgeek 11y ago

The problem is, those existing systems were there for a reason. And I'd guess the reason usually was that the previous, less formal system also had their weaknesses. Most notably, they are very probably very subjective. IMHO performance ratings and deciding who gets a promotion simply are very hard problems, and no good solutions. So people are always dissatisfied with the current approach, and so add rules and layer to fix it. Eventually, the system becomes unworkable and bloated, and are stripped down to the basics again. Which we currently observe. But I wouldn't be surprised at all if the cycle continues, and more rules and formal processes will be added.

yuhong 11y ago
I do think trying to quantify performance into a number is flawed for anything other than factory work and the like.
- perlgeek 11y ago
  Yes. And the dilemma is that there has to be one number (the one on the pay check), and it'd better correlate with performance in some way.
  - yuhong 11y ago
    I think a better approach is to give the same salary to an entire team.
    - eitally 11y ago
      Why do you think this is a good idea?
    - solomatov 11y ago
      The more creative the work, the higher the difference in performance. It seems to, it's unfair that people who do substantially more, are paid the same amount of money as underperformers.

kfk 11y ago

I think it's good to sit and talk about performance and work every now and then, I don't think it's good to write things down and put them into a system for years. Also, I think it's horrible to have things like OLR where "leaders" sit and talk about you and how well you do (from below to above expectations). The thing is, most managers are bad, especially the middle ones, and their opinions about their reports tend to be bad and flawed. Managers get overly excited about extroverted people and the introverted struggle because they don't show "leadership skills". However, in many fields (i.e. mine in finance), you need a high % of introverted people because they are the ones that usually perform the work, while extroverted spend time socializing and "networking". This whole situation creates a continuous incentive to game the system - instead of doing your job, you spend time working out ways to game the reviews (there are many, the classic is when you are indispensable for some mission critical tasks). The result of this is that you will get politicians at the top, those that game better the system, instead than results. It's not easy, but definitely performance reviews and OLRs are not the answer. I think 1:1's are a good idea, because they happen more frequently. As for feedback from others than your boss, I think it would make much more sense if that was on the whole team/department, rather than on single individuals (like in OLRs).

zekevermillion 11y ago
The extrovert vs. introvert distinction explains a lot. I'm an introvert, and have always hated reviews, even when I know it's going to be a good review. The best review in my opinion is when I get no response at all -- that's when I know I've done my job, and it just worked. Extroverts seem to enjoy getting "feedback" even when it doesn't matter -- e.g., when there's no raise in it, and there's also no chance of being fired. Unfortunately in most companies, compensation (even that which is nominally based on performance) is not based on performance, but rather on the market. An accurate comp-linked review would go something like: "we decided to give you a 1% raise this year b/c we think that is the minimum necessary to keep you in this job; also we used up our budget to hire this other person, and she needs 10% more than you're making to convince her to leave her other job."
- ild 11y ago

Intermernet 11y ago

Performance reviews are a form-letter attempt at managerial automation. They seem to suggest that a manager should only be monitoring the performance of their team at certain points in the year. A good manager (a rare breed) knows to a fairly certain degree how their team is performing at any time based on the output from the team. If they don't know this, how are they meant to manage? I flat out refused to do performance reviews at a previous position. If the company has an issue with my performance, then HR can talk to me directly. If the company wanted to fire me for not filling out their performance review, then I didn't really want to work for that company. They never mentioned it, and I left the company amicably.

dbg31415 11y ago

Without a formal time for review, how will they build a formal time for pay evaluation and bonuses? I can't help but think this is like the latest cell phone craze... yes, your bill drops down slightly, but you are going to have to pay for your next phone and when you do your bill will actually increase. Suckers. Stack Ranking was and is fucking horrible. And getting rid of that for honest 360 reviews should be the goal. Reviews from your peers with project completion, reviews from your manager with year completion... those are both important and can help you grow. Not to mention get a raise.

VLM 11y ago

“This person is ready for promotion today.” I would wager the percentage of employees working at a firm that still participates in promotion culture is low and dropping. In other words "ready for a promotion" for most people in 2015 means they're going to quit soon and get a higher level position at another firm. I'm sure there still exist individual companies and maybe unusual and rare fields where promotion culture still exists, but it seems very rare and unusual. Meanwhile in most areas outside SV/NYC computer programming during bubbles, there is a large excess of qualified workers vs positions. For example in K12 education the state-U system graduates twice as many teachers as there are open positions for all new and past grads to apply for. It doesn't matter if they are in the top half that got hired because they are the best, or the most politically untouchable, or the most attractive, a year after hiring at review time they're still going to be that top half. So there's nobody to give a hard time to in a review, because that lower half or lower 3/4 or whatever in any professional field are already working at Starbucks or as a waiter or bartender. The world has "needs improvement" kindergarten teachers, but they don't work at your neighborhood school, they never got a job in teaching and they serve you beer at the bar, for example. The economy is in permanent contraction and the general population and higher-ed sector are in permanent expansion for many years now, outside bubble segments of the economy, so hiring one of the bad workers strongly reflects on a managers interview skills, making it even harder to admit anyone is below "excellent". Meanwhile increased labor mobility for the few that are employed means a review a couple years back, is a couple employers ago, so it is inaccessible and irrelevant, so its not like the wasted effort today will pay off in ten years. Maybe a review in 1950 would come in handy at the same employer in 1960 to select a new director, but the firm has no loyalty to the employee and therefore the employee has no loyalty to the firm, for at least three human generations now, so its been culturally bred out of us. So the folks at the top of a team will interpret an attaboy as "start applying for higher level jobs elsewhere" and economic realities mean you don't have truly low performance employees to crack the whip upon, making the whole dog and pony show a waste of time for all involved. This leads to an attitude of just pencil whip the process to get the meaningless metric of "accomplished reviews" up to 100% and be done with it. Once the folks who've been pencil whipping it for their whole careers make it up high enough, its finally politically possible to discuss getting rid of the waste of time. So here we finally are.

11y ago

solomatov 11y ago

It's very easy to criticize performance reviews. Does anybody know about more sensible alternative to this system?

greggman 11y ago
I'm curious too. One company I worked at had me privately rank coworkers. No Idea if it worked. It seemed like it might identify top and bottom performers?

lectrick 11y ago

In a creative job (such as programming) it's extraordinarily difficult to evaluate how valuable someone is, but I would suggest some sort of human valuation ("you know it when you see it"), a LIGHT metric used only as a data point (goals and whatnot, any measurable impact on bottom line, or website performance, or some other numbers), and some means of simply measuring how much someone cares about their job, company and career. I think I can tell how much the people I work with care about their job, company and career, I think I could trust most of them to know that about me, and I think that's a good place to start. Metrics-wise perhaps a "spread spectrum" approach that covers so many things that most signs of being "productive" would end up at least reflecting in the numbers, yet would prevent "gaming" of the numbers since so many things are measured. Used only to inform and not base hard decisions on. This would catch both people who care about their job but for whatever reason aren't producing anything measurable, as well as people who have produced but perhaps could produce even more if they were more engaged.

solipsism 11y ago

The yearly performance evaluation is a joke. Any company which respects its employees as humans and which takes seriously its responsibility to its shareholders will promote the ongoing evaluation of employees based on close 1-on-1 relationships between manager and direct report. And the quality of the performance evaluation at any company that does this is only as good as the quality of those managers. If all your managers are people who are out to make themselves look good, or out to keep their own jobs safe, no system could possibly get the most out of your employees. If your managers are motivated to care about the company and to make its goals their own, and if you provide the proper training and evaluation of their management to ensure they know how to manage the right way, you have a chance.

markbnj 11y ago

I'm happy to see this become a topic of conversation. The formal performance review is emblematic of the key difference between small and large organizations. A small organization will usually rely on ad hoc information gathered from the process of daily interaction and observed results. Opinions form, are solidified, and ultimately business decisions are made as informed by these ad hoc impressions and the resulting attitudes. For the best small organizations these approaches work very well. Managers have enough personal oversight of the company's operations to feel like they have a good handle on what's happening. Once an organization grows to a certain size management begins to lose that level of personal oversight. No one person or even a small group of persons can comprehend the scope of the company's operations. More levels of hierarchy have been introduced; there are more managers responsible for more people. Now it is probably possible that those same ad hoc approaches that worked in the small organization could continue to work at some unit level of the larger organization. Managers could maintain personal oversight of their unit, the observed results could percolate up the hierarchy, decisions could be made based on those results. Unfortunately, the insecurities of higher level managers will not usually allow this to happen. Lacking that personal oversight at the operational level they attempt to compensate for the risk of a bad outcome by developing rules which ensure a good outcome. This happens so reliably as organizations grow that the emergence of rules is nearly synonymous with the concept of growth. However, and I think this is the key point, it seems to me that developing rules to ensure that an organization is managed to a good outcome is a little like trying to develop rules that ensure a baseball team will win a majority of its games in a given season. Rules of this sort actually cast in concrete a specific view of and response to a set of conditions that exist or existed up to the time when the rules were created. They stifle the ability of the organization to react to changes in conditions that may not have been anticipated. And for whatever reason, probably due to essential human nature, they seem nearly inevitable once an organization reaches a certain size. It would seem to make sense to act more like the baseball team: acquire the absolute best people, give them the absolute best training and resources, make the mission absolutely clear, and then reward or punish based on results. But to take such an approach requires waiting for those results to come in before you can know whether you were successful. Which is, interestingly, how military organizations have to handle field commands. When battle looms the hierarchy (and not inconsequentially the troops) has to put its faith in the people in command on the ground, whom they have put in place, trained, and supported. The tides of battle change too swiftly, and the outcome of poor or slow decision making is so brutal, that a top-heavy approach to governance simply doesn't work. In some ways that is the essential story of the way the war in Vietnam was fought, for example, as contrasted with previous conflicts.