93

I had this manager that frequently brought up how team members performed compared to each other during one-on-ones or performance reviews. E.g. "A could do this task in half an hour you took three hours on" or "I can not give you a better performance review because B delivered much more than you did and I am giving them an average review already."

Now the tone on the examples aside how acceptable is it to bring up other team members while discussing performance?

  • 1
    Comments are not for extended discussion; this conversation has been moved to chat. – Jane S Jan 17 at 5:00

11 Answers 11

148

It's not.

If I'm evaluating your performance then Bob and Alice have no bearing on the review. Now truthfully I'm human and I may think of comparisons to each other (and to other people I've worked with over the years) but I need to deliver my evaluation about you in as unbiased a form as I can.

It's the same principle that applies when I reprimand you and you bring up Bob and Alice.

Bob and Alice aren't there. It's just the two of us.

  • 6
    Could you elaborate why you consider a work environment a place for a neutral, individual evaluation rather than one that tries to optimize its productivity by having everyone deliver the best they can (including replacing them, if someone else can even deliver better results)? (I like your statements, I can just see them invite the usual "We're a business, not a charity (or, in this case, free self-improvement service)." response and wonder why that wouldn't apply.) – O. R. Mapper Jan 14 at 8:19
  • 81
    @O.R.Mapper Comparison between things usually requires something measurable. Getting something measurable means metrics. Having metrics mean you will end up with employees good at generating metrics, and not thinking about their work. If you then FIRE employees with low metrics, then the goal of the employees will be to make sure someone else gets lower score than them. That's much easier than actually do consistently excellent work all the time. This is called STACK RANKING. It's very bad. Read up on Microsoft and Steve Balmer. – Nelson Jan 14 at 9:00
  • 2
    Performance metrics in my experience are also not an effective way of comparing employees with each other, rather they are a measure of where they are compared to where they should be. – DoctorPenguin Jan 14 at 13:18
  • 2
    I think openly comparing performance could be tactless and hurt morale, but I don't see how your statement is at all true. Performance is relative, Bob and Alice are reference points to establish what an employee can be expected to accomplish. How they perform is absolutely salient to a performance review. – John K Jan 14 at 18:38
  • 5
    @bruglesco It also assumes that each employee is an island. Teams I've worked on have often either been made up of a variety highly specialised functions (so you are really comparing apples to oranges and is why performance metrics are IME a terrible way of comparing staff with one another), or comprise a team which support each other accomplishing each others tasks. I've found that the "highest performers" can be the ones who take all of the credit for work that other people were heavily involved in and the "lowest performers" were those not assertive enough to tell these people "no". – DoctorPenguin Jan 15 at 11:08
95

This is a MAJOR red flag and an indication of toxic management.

I would seriously consider either completely ignoring the fact that this manager is trying to divide and conquer or find another job.

  • 85
    Don't attribute to malice what can be adequately explained by incompetence. Most managers think along the same lines, they're just not stupid/tactless enough to say it to the report's face! – jpatokal Jan 14 at 5:14
  • 10
    Thank you for adding the mandatory "find another job" answer. – Chris Jan 14 at 8:39
  • 5
    @vlaz As an aside, it may matter if you want to try to change the behaviour. A manager doing this maliciously will probably not be convinced to stop doing it; a manager who doesn't realise the harm they're doing can probably be "corrected" with a single, quiet conversation. – Lightness Races with Monica Jan 14 at 13:07
  • 5
    Toxicity does not require intent – Dancrumb Jan 14 at 21:58
  • 5
    @Džuris yes. To follow through your scenario: Now I am annoyed that my manager said that to me, I am resenting Alice that she made me look bad and if Alice makes a mistake I'll make a mental note of it. Next time the manager pulls me up, bam! I say "Alice did that in 30 minutes because she didnt follow X procedure". Alice is of course told about this by our manager and now she is watching me. The manager is happy that everyone is watching each other, short term productivity has increased (long term is a different story). You get my point about how this could lead to a toxic workplace. – solarflare Jan 15 at 21:41
47

There is a reasonably well-known practice called stack ranking, where businesses explicitly and intentionally compare employees, in an attempt to get rid of a small percentage of unproductive ones, and reward a small percentage of exceptional ones.

Your manager's statement “I can not give you a better performance review because B delivered much more than you did and I am giving them an average review already” sounds a bit stack-ranking-esque to me, and might indicate that your company is doing that.

Stack ranking is often criticised as being really destructive for morale. That's certainly how I felt in a job many years ago, where I believed I'd put in a year's worth of really good effort and done everything we'd agreed, only to be told I couldn't get a top ranking (and therefore a higher pay rise) because another (very deserving) team member already was, and there was only one top spot allocated on the curve.

Of course, this might not be a considered company-wide approach to performance measurement — I don't think that employees are traditionally told about other specific employees' ranks in a stack ranking system, and it would usually be reasonable to expect discussions in one-on-ones to remain reasonably private. If it is just your manager doing this, and only doing it in private one-on-one conversations, that doesn't sound like particularly good leadership.

However — if the company actually does want to directly compare employee performance, maybe to encourage competition and/or collaboration to drive improvements overall, there's no reason why they can't do that. Sports teams (rowing is an especially good example) have very intense metric-based competition between team members, who also have to very closely collaborate to succeed.

Obviously it's difficult to translate that to knowledge work, where the performance metrics are much vaguer, but the problem of a team with both internal competition and internal collaboration is the same, and some comparison of team members with each other — hopefully in a less negative and divisive way than you've described — could be a part of a successful approach to solving it.

  • 36
    Stack ranking is claimed to be responsible for a “lost decade” at Microsoft. – gnasher729 Jan 14 at 10:19
  • 26
    @gnasher729: even stack ranking can't be blamed entirely for Vista. – Paul D. Waite Jan 14 at 10:38
  • 4
    Even outside stack-ranking (which is quite an extreme thing), most places I've worked in did their reviews and then did a 'calibration' exercise where, essentially, teams with more high performers than the average were talked down a bit (and possibly likewise for low performers being talked up?) Thus, this manager may be internally thinking like the upcoming 'calibration meeting' will demand. – Ralph Bolton Jan 14 at 15:52
  • 3
    @gnasher729 that was attributed more to the fact that they did it per-team instead of company-wide, so a very good person in an extremely good team would be fired, and the second-worst person in a terrible team would keep their job. – Robert Grant Jan 14 at 16:58
  • 2
    @gnasher729 And also for a "cult-like" culture at Facebook. Link – code_dredd Jan 14 at 19:05
27

There are some good answers already, but they each seem to capture only part of what is quite a complex answer.

There are many reasons why it might not be a good idea to directly compare staff members in a one-to-one performance review:

  • An employee's performance should be judged against the expectations that have been set for them, not anyone else, which will vary with their experience, salary, goals, etc.
  • Good team members should not be punished because another team member is outstanding.
  • Different team members bring different (ideally complementary) skills and are probably set different objectives.
  • Setting team members against one another is likely to foster a vindictive atmosphere.
  • Whatever metric is used for the comparison (e.g. time taken to get work finished) may be "gamed" and become the metric that team members prioritize, at the expense of other considerations (e.g. work quality).
  • Restricting the proportion of employees that can get a good rating (however that is measured) is demoralizing and simply not in line with reality.

However, there are also reasons why it can sometimes be a good idea to compare staff directly:

  • High-performing staff can be used as examples for others to learn from. "Watch how Alice solves problem X" is easier advice to follow than "do better".
  • The actions of other staff are concrete and undeniable, which can be useful for accountability. For example, "You just told me it's not possible to do that work in less than five days, but Bob did that exact same job in three days".
  • There is no Platonic ideal for what a good worker (in whatever job/career) looks like. If I want to employ a software developer, what should I expect from them, and how much should I pay them? The only rational answer is to compare them to others (though to be fair, it might be tactful to be circumspect about making that comparison).

In addition, discussing or comparing to other staff in a one-to-one or performance review can be desirable for other reasons:

  • If someone is being considered (or has longer-term ambitions) for a promotion into a leadership or management role, they are going to need to be able to assess other people's performance. Demonstrating how to do that in a one-to-one might be part of a longer-term development strategy for that employee.
    • (Anecdote: one of my team once told me he wanted to be a Development Manager in 3-5 years because he wanted "to be able to tell people what to do". I made sure to involve him in the kind of discussions a manager would need to think about, long before he was likely to get that promotion. He no longer works for me, but I can only hope that if he has been promoted since, he has a better understanding of what a manager is and isn't just "telling people what to do"!).

Finally, in response to the inevitable (hysterical) "quit your job!!!" answer, consider how likely it is that management in your next company will be perfect. If a manager telling you "Alice is doing a better job than you, but you're doing better than Bob" is the worst thing that happens at a company, it might still be an OK place to work.

In summary, this is a complex question that resists simple soundbite answers. Depending on context and how it is done, it may be a poor choice on the manager's part, or there may be a good reason for it.

  • 2
    +1 for different team members bring different skills, that's what makes this mostly useless IMO for any team that does difficult work. you may well be comparing one members strength to another's weakness, doesn't mean both aren't valuable on the whole – aw04 Jan 15 at 15:06
  • 3
    Nitpicking a bit on one example: depending on the industry, "You just told me it's not possible to do that work in less than five days, but Bob did that exact same job in three days" can say more about a manager's inexperience than anything else, so it might not be a great example. Software development in particular is notorious for completing a 3 day tasks in 20 minutes or 20 minute tasks in 3 days. – Morgen Jan 16 at 5:40
  • 2
    @Morgen and it gets even worse when the person doing the estimation isn't doing the work, because the other person is highly unlikely to do "the same task", but will end up implementing things differently. – Erik Jan 16 at 10:03
  • It was intended as an industry agnostic example. This is the workplace, not software engineering. Though even then, while there are different ways to achieve the same task in software, there are better and worse ways to do it, and there are better and worse staff. Comparisons and stack ranking must not be abused to crush individuality and different valid approaches, but some staff (and some software developers) are indeed better than others, so let's not pretend otherwise. – BittermanAndy Jan 16 at 10:13
  • @BittermanAndy software is simply an easy example, because it's especially notorious for wildly inconsistent estimates in both directions. General contractors, repair technicians in all fields from IT to plumbing, writers, lawmakers, and essentially every profession that touches the unknown has issues estimating to a greater or lesser degree. Often times an order of magnitude difference can be traced to a missing bit of knowledge and correcting this is more important than making value judgments about the people involved based on this particular metric. – Morgen Jan 16 at 16:35
11

There was just an article in the New Yorker about the two top engineers at Google and how they complemented each other. They ALWAYS paired together. (Sorry if I get some of this wrong, it's from memory)

The article claimed that they were the prime example of how one great programmer could do the equivalent of 10 normal programmers and together saved the entire company repeatedly solving problems that nobody else could.

It sounded from the article as though one of them generally typed and the other often sat there with his feet up on the desk and thought.

Imagine the effect trying to pit them against each other or force them to identify who did what work so they could promote/adjust pay/fire one or the other... it would be absolutely destructive. It would have ended the company.

So I'd say that if it was acceptable by your company, you should try to correct that or leave. You might stick around and slog through it like BittermanAndy suggests, but it's not a practice that shows any understanding of software development teamwork so be ware.

I suppose that it's also similar to rating your basketball team only on baskets per player because assists don't score anything. Would it be better to get rid of players who don't score as much or figure out that you are scoring wrong? And making this mistake publicly in front of the whole team when they can all easily see how wrong you are???

  • 4
    I really enjoyed reading this answer and it offered good perspective...are you able to link the article? I'd enjoy reading about it! – Jessica Tiberio Jan 14 at 21:55
  • 2
    @JessicaTiberio Feel free to correct my post if I mis-remembered any of this: newyorker.com/magazine/2018/12/10/… -- and yeah I found it a really good read. – Bill K Jan 15 at 17:39
10

I second @solarflare's answer, but I'd like to add another reason.

Do consider that maybe the only inappropriate aspect is giving names explicitly. Rephrasing to "I can reasonably tell the task you took 3 hours could have been done in 1 hour" or "I cannot give you a better review, because of the standards I follow, which also normalizes every employee's review". Would that sound bad to you?

My point is that I believe comparing team members explicitly is poor ethics, but for all practical purposes, the same message could have been sent without the comparison, which would have been made either way.

On the plus side, you are given the opportunity to dispute the comparisons, saying things such as "A took less time because his task was actually simpler" or "I actually helped B a lot in delivering the tasks that you are attributing to him". Likewise, these answers do lack professionalism, but so does the person evaluating you.

I do believe this is not a reason to find another job, but coaching managers into giving proper feedback is pretty much a HR responsibility, maybe seek the HR department with some constructive suggestions.

  • I'm pretty sure throwing your teammates under the bus (yeah, C only took an hour but that's because they did a bad job) is going to be awful for morale.... And then you also have to start talking about what you're paying people (my architect only took an hour, but you pay them 3x what you pay me, so we are both getting an A for dollars/hour), and no company wants you to start comparing how much you make.... – user3067860 Jan 14 at 14:23
  • @user3067860 : Hence why I said it's poor practice of the manager to begin with. And even the examples I gave (softer than yours) would be already very poor attitude. But then again, good professionalism may be unsustainable in when dealing with very unprofessional people. And if you are misjudged due to an unfair comparison and you have your change to speak up against it, failing to do so means agreeing to the comparison. – Mefitico Jan 14 at 15:04
5

You did not have a manager. You had an excuse for a manager.

The reason that your "manager" failed as a manager goes much deeper than the simple fact of telling you that Bob got the raise, and you didn't, because Bob made X widgets last year/quarter/month, and you didn't. That is merely the cherry at the top of the bonehead sundae.

Making performance review a competitive affair was itself an exercise in idiocy. No surer recipe for distrust, resentment, and in-fighting was ever created than turning employee compensation into a zero-sum game.

We can go further. Performance review is itself inherently flawed. It assumes that employees are in full control of their performance, when the truth is that employees typically control only a few of the factors that influence performance. In practice, the only reliable result of performance review is to cripple the morale of employees who receive bad reviews.

On top of all this, your manager's performance improvement toolbox seems to have sticks and carrots—with emphasis on the former, I'll wager—and nothing else. Any idiot can hand out rewards and punishments. That is not a skill set that merits anything above the minimum wage. When management really cares about improving performance (instead of caring only about looking like they're doing something), they look into what the better employee is doing differently, and try to implement that employee's methods, techniques, and practices on a wider scale, and make sure that every employee receives the necessary support for their job.

  • "Performance review is itself inherently flawed. It assumes that employees are in full control of their performance, when the truth is that employees typically control only a few of the factors that influence performance." That deserves highlighting. – catfood Jan 16 at 19:58
3

It might be a good thing if it is presented in a positive way : two employers ago I have been working in a project team and every week somebody was elected by the project manager as the "employee of the week", meaning the person who has done something exceptionally good. There have been cases where two people had done something exceptionally good and the project manager exceptionally elected both.

But saying something negative in public is a show-stopper: such a thing should never happen.

0

In order for it to be acceptable it is necessary (but not enough) that it is a complete and accurate comparison. Which is quite a difficult thing to do. Your case shows how sloppy the comparison is with just a cursory review. It's missing a crucial element : paychecks. A is not under performing if she takes twice the time than B but gets a third of the pay. She might actually deserve a rise.

In general you won't see managers revealing your co-workers salaries. It's even illegal in many places.

Hence you are unlikely to ever see a proper comparison of team members performances. This is not to say that if the comparison was accurate and complete it would be acceptable but I have nothing to add there compared to other answers.

-1

The act of comparing coworkers is entirely valid and a useful tool.

The issue is not that you're being compared to your peers, but that you aren't being compared against the expectations that (should) have been set with you. For example, working on commission is an example where your expectations and compensation is tied explicitly to your performance relative to your peers.

The only way that performance reviews go well is if you have a regular expectations/goal setting cadence. Then you get to be reviewed against those goals, which may or may not inlcude comparisons to your coworkers.

  • 14
    They're not "delusional", they're based on fact and experience. Sales is pretty much the only place that open competition of this form is expected, and that's an industry which already attracts a very specific kind of character and has success at least partially driven by luck/market conditions. If you were to apply your approach to a software development team you'd find yourself with an abysmal environment very, very quickly. Real-world history confirms that this is so (as discussed in some of the other answers). – Lightness Races with Monica Jan 14 at 13:09
  • 3
    @LightnessRacesinOrbit just to add onto that - a company I worked for didn't exactly do this but would regularly praise one team or another for accomplishments. This is fine but the problem was the regularity - some teams wouldn't get a mention simply because they hadn't done when "praise time" came around, they've been stuck fixing bugs or designing/developing something that would come in soon. Yet people who recently did a successful release would get a mention. It was pretty annoying for those involved even if it's not quite the level of constant comparison. So, yes - that would be worse. – VLAZ Jan 14 at 14:21
  • 1
    @vlaz: Yep have experienced this too. It's very demoralising. I almost quit over it. Ironically the same problem exists in reverse on occasion: praising the least skilled workers as a means of encouraging them, which sounds great on paper but if you forget to ever praise those actually accomplishing things in the meantime then you're risking disenfranchising your best workers. – Lightness Races with Monica Jan 14 at 14:37
  • 1
    @AdamMartin That part is certainly a valid observation. – Lightness Races with Monica Jan 14 at 16:00
  • 6
    I'm surprised this has so many downvotes. As an answer I think it's incomplete, but it's not wrong. – BittermanAndy Jan 14 at 16:50
-1

The comparison itself is very similar to the idea of the of normal distribution or bell curve in performance in assessment, which is perfectly fine and valid. It is to be expected that in your company you'll always find non-performers, average performers and top performers creating an unequal distribution of performance. This is not an invention of managers; this is just a natural phenomenon. In your example B is the average mediocre performer. If A underperforms according to the expectation set by a dynamic median of the curve then it is perfectly reasonable to make A aware of that fact as part of your plan to improve performance.

However, it is not useful at all to openly compare performances with coworkers present, because that's irrelevant to the improvement of the team. Team members are not supposed to compete against each other which is the implication of the open comparison. Team members are supposed to compete together for a common goal.

You'll always get non-performers, average performers and top performers, even if everybody improves substantially, hence a dynamic normal distribution of performance. But the goal of this assessment should be to improve the median performance of the team, not to make everybody a top performer, which is nonsensical, or punish them for being mediocre or non-performers.

  • 2
    I think saying always here is a bit too dramatic. You do not generally get a normal distribution on anything until you have n>20 and the average team tends to be much smaller than that. At least a team you can find time to compare people openly like that. – Victor S Jan 15 at 20:33
  • @VictorS You are right, of course you will never get a *perfect normal distribution of anything unless n is infinite among other things . That's not my point at all. According to the situation given by the original post , the performance of A is being compared with B who represents the median: "I am giving [B] an average review". Please correct me if I am wrong. Even if "n" remains unknown that doesn't matter since the assessment follows the *idea of Normal Distribution, or at least an unequal/uneven distribution with a median. – user97962 Jan 15 at 21:32
  • Of course, if you pick a person as a median within your team you can rank people above and below them. Calling it bell-curve-like gives it a misleading legitimacy though. For all you know, all team members may be performing well above median within a larger, relevant sample. So you end up ranking and penalizing someone for not achieving over your median while under normal circumstances (e.g. goal based evaluation) they would pass with flying colors – Victor S Jan 15 at 21:42
  • @Victor S If you are not comparing team members then you are comparing teams. Still the same idea. If you want to improve the mean performance of the teams then you want to improve the performance of each team. If you want to improve the performance of a team then you might want to evaluate how team members are performing. And you'll always find a distribution (no clones). That doesn't meant that you can justify or legitimize any counterproductive measures upon that, such as openly discussing scores with other team members present. Not such an implication – user97962 Jan 15 at 22:30
  • Productivity bell curves don't happen everywhere. In many fields, if there's a bell curve of ability, companies hire more or less from the right side of the curve, and so it's likely that the majority is below mean performance. – David Thornley Jan 17 at 16:20

protected by Jane S Jan 17 at 4:59

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.