Ask HN: Why are annual reviews so broken?
It's seems to me that if companies drew a straight line between here's what you need to do to get this much of a raise or this much of a bonus, they could get alot more out of each employee. why waste everyones time and effort for stuff that doesn't matter?
85 comments
[ 3.1 ms ] story [ 152 ms ] threadSome people want it straight, do this and get this, but we have gone to a way more mild approach where it's no longer an 'annual review' but more a 'wellbeing meeting' because some people didn't like the format and felt that they didn't want to sit and listen to being critised and that the metrics were unfair or not applicable to them.
So now what we have is much worse, last year, 30% of the team got a pay rise as is company policy, however we do not know how this was dedicated and most of us have barely spoken to those that do as our direct line manager was not involved in the decision.
Personally I hate them, I just want to be left alone.
Instead it should be a conversation around the two way relationship where both sides at least have the power to exit the relationship. Both sides should approach it as a conversation about the relationship and things that work / don't work from both sides.
The first is best done by peers & management anonymously, the second by a fairly objective public rubric and the last is purely market driven (no matter how much companies wish it weren’t).
Collecting all the information the same way and using the data collected cross purpose leads to all of them being done poorly.
Why would companies wish it weren't? It lets them underpay people because "the market" broadly underpays them.
I've talked about this dynamic with leaders like that, who opposed salary transparency because they underpaid people who they knew the market would undervalue, where the actual contribution of the person in question is irrelevant. As they put it (grossly), if you're a minority that would be paid less elsewhere, I'm going to pay you less here because I can. Not because you're a minority, but just because that's the way the market works.
I've been in situations where the "market forces" are stupid enough that I've been paid more than someone nominally one level up from me (senior eng. vs junior). It's immoral as hell but what are you going to do? The company is always going to pay every worker as little as they can, and not a cent more.
Is it illegal? No. Immoral? I'd argue yes, if the only reason you get to pay them less is that the broader market is biased against them for reasons outside of their control.
Edit: see also the recent stories about home values varying wildly based on whether the current owner is white or black. This is based on perceived 'market value', is this fair? https://www.cbsnews.com/news/freddie-mac-home-appraisal-hous...
However, having evidence that that is the reason such that we can be confident it is that, rather than assuming it, is difficult.
You might say that it amounts to the same thing, but attributing the problem to the correct cause is important.
It's important because it means that if the correct cause is bias, it's actually bias in every potential employer somehow, but how do you know? How do you know it's not something with the applicant or the application?
Saying it's just a biased market means you don't even think of other things it could be, such as a badly presented CV or a missing key skill, or a difficult interviewing manner. Those are more likely to have every future employer reject you than something insidious. It's unlikely that every potential employer has such a large pool of candidates that bias is the thing that means you don't get a decent enough offer to even beat your current salary.
But if it is a cabal of horrible biased people, then that's illegal and should be challenged as such. It's horribly unfair, but that's probably the best course.
Suffice it to say he started saying he was sorry but despite his efforts he could not convince the higher ups how stupid an idea that was.
I don't know the solution, other than just have general expectations about assigned work and commitments as those commitments come up and change. Rather than having some performance conversation that's out of sync with actual project deliverables.
That's because their boss (or their boss's boss or whoever) is telling them to "get it done"/do that.
Seems like a lot of effort for the typical 3% raise most people get more often than not lol
If you work at a place for 10 years, is it really realistic to expect to get a $10k/yr+ raise every year? Goals or no goals.
Inflation is at 6.4 percent right now. If you made 100k that would take you up to six and a half just for treading water, if you make 200k then and get 10k you are being paid less than last year.
In my experience, whether inflation is 2% or 6% or 8%, your employer doesn't care. You have next to no leverage to demand a constant static cost-of-living-pegged-1:1-to-inflation raise.
yeah, if inflation is 8% and you don't get an 8% raise you effectively get a paycut. What are you going to do, strongarm your employer and demand they give you at least a "CPI trailing 12 months" raise every 12 months? not realistic in my opinion/experience
Anything over $160k/yr, I feel like you are subscribing yourself to be "the go to person" for everything and a lot is expected from you.
More like it for HR/managers to justify their presense by "doing something"
If you aren't getting a wage increase above the annual inflation rate each year then not only you're not getting a raise, you're actually getting paid less for the same work (or, rather, for even more work, as your responsibilities will increase with the seniority).
For 100K wage that's like getting a 7K raise atm.
You're not wrong but I don't think it's realistic to expect every single year your employer to give you a "1:1 with CPI" raise.
If anything, maybe 0% one year or 2 years and then maybe a 10-15% next year (if you worked on some really key deliverable projects where it can be justified).
Is that your experience?
For the last decade tech has been hiring like crazy and jump jobs has been easy. I can't speak for how that will change going forward, but it does concern me.
Needless to say that rarely happens, and even when it does there are often no available followup assignments in the same area to reinforce the knowledge, so you move on to something else and forget 80% of what you learned.
Your relationship with employees is through salary, that’s it. Not the comfy chair, the firewall rules, pool table, or “agile” training… but salary.
Let them talk salary and you will find they provide healthier boundaries. (Your comment gave me no indication of your opinion, I was just adding.)
In sales it’s fairly simple to set clear goals that can be evaluated in the review. I think for a lot of other people goals constantly change during the year. It’s hard to evaluate that. It could maybe work if management constantly gave you clear and achievable goals for the month but that would require a lot of effort by management.
In the end annual reviews are BS for most people. “Nice job last year. Keep sound what you did. We don’t have money for raises.”.
1) there isn't very good definition of what most jobs are.
2) the definition isn't pulled through all your human capital management systems (recruiting, onboarding, career management, performance management, and off boarding).
So, I'm working on building a product that pulls 1 + 2 together.
But the positive-sum scenario would be: employees will be more productive if expectations are clear.
Here's an easy example. If you were going to get rid of someone in 4 months after their review, would you give them a raise? Of course not, you'd allocate that funds to someone else.
That's why it's hard. The business has to judge which people they are going to focus on keeping.
A normal interview these days is 38 minutes (45 minutes minus 2 minutes introductions and 5 minutes for interviewee questions at the end), times 5 or 6.
Some leetcode, some Jedi (cultural fit), a design sketching or 2. People just go hard on leetcode, grokking system design, and memorize a few situational answers in STAR format and they're good to go.
How good will they be at identifying why the Kafka consumer is lagging at midnight? Who the hell knows?
So a broken system brings in thousands of new employees, now you need to define success to all of them at the same time, and figure out how fast you can get rid of those who can master leetcode but not deliverable production quality code
Because they can get more out of the high-performers by not giving them a clear threshold (which is why they are called 'over-achievers') and if the mid to low performers met the threshold then the company would have to pay out a lot more in salary.
If you run a company you know this, but for the uninitiated: there’s a lot of work to fire someone in the US. Technically most employees are at-will, however companies need to have non-sueable reasons for terminating employment.
If they fire you because of race, sex, or some other protected class, that’s illegal. The employee can file for unemployment and that makes your company’s rates go up.
So before terminating someone, companies get a lot of written paperwork in order. This is typically called a performance improvement plan (PIP). That usually starts at the Annual performance review - they’ll say you have poor performance in specific areas and write it down in traceable documents.
The employee tries to “improve” sometimes, but if the company is adamant on terminating someone it’s really just a matter of time.
Their goal is to have a strong case (aka paperwork) demonstrating poor job performance so that when you sue (or file for unemployment) they can push back with a 6 month history of documented issues.
I don't particularly like managing people, and it's easy to slip into a lazy management style where you avoid confrontation, but the truth is that I've found people respect honesty and a chance to course correct.
That's a management fail, to be honest. The first time someone hears they aren't keeping up should not be because you started paperwork on them.
Seriously. That's alarmingly bad management.
For example, my annual reviews aren't so much about setting and discussing goals as it is about how the year actually went, both from a professional and personal standpoint. We talk about projects we finished or didn't, how the company and team is doing overall, and big strategic decisions made by execs that we don't really have control over but impact us in one way or another. Goals are generally set pretty frequently and are achievable. I have 1on1 with my leads every 2 weeks and we exchange pretty blunt but constructive feedback on how we are doing.
Even in clear cases of misconduct it's common enough to simply terminate.
The assumption is they'll get more effort out of labor if they're uncertain about the outcome. No matter how valuable an employee is the company must always maintain a MAD type posture that no employee is too valuable to lose, because if that were true and the employees knew they would take the company over effectively.
Part of what the annual review is is a reminder that your job is not an entitlement. It is not supposed to be pleasant for the average worker.
From the employee's perspective, there's usually little reason to take annual reviews seriously. The fact that they are annual and not more frequent subtly communicates that any negative (or even positive) feedback really is not that important, if at all, and savvy employees who care about maximizing their earnings should consider whether it's worth investing effort into "improving" when said improvement is unlikely to translate into a bigger paycheck or a fancier title. Some companies try to counter this by having employees self-review, but this backfires in a number of ways. For one, employees are incentivized to either exaggerate or downplay their strengths and shortcomings. "Always room for improvement", right? If an employee legitimately can't think of anything they need to improve upon, they're liable to make something up just so they don't stick out like a sore thumb. One way that this system does work out for employers is that it can encourage less savvy employees to work harder than they did before but for the same pay, but this may also cause burnout and higher turnover when these peons realize their station at the company and move on to better things.
In rare cases, annual reviews can be used in an effort to get rid of poor-performing employees. This is not as common as people think, and is even less common the bigger the company gets. Businesses want to maximize productivity, but that doesn't mean they care enough to outright fire the mediocre or even lazy employees. Mediocrity can actually work in the favor of the business because, if such mediocrity is permitted, they can hang on to employees longer than if they demanded the kind of extraordinary excellence they claim to in job ads. "We only hire the best." No, you don't.
Many people like to do things outside of that box, or hate being so constrained.
Companies also struggle to quantify the value of some of the things people do.
For example
- Employee 1: Sales rep, hits quota, shows up late, leaves early
- Employee 2: Sales rep, behind on quota, but everyone in her pod is crushing because she's spending extra reps helping them be better at their job.
It's likely that employee #2 is in the wrong role. But if you're a mega-corp manager, all you see is that she's behind on quota, and her annual review reflects that. Let's put her on a pip and find some more people like Employee 1.
It could be that Employee #2 isn't motivated by more money.
Telling her "Do more sales or you're fired" could end up being pretty bad for you.
You may end up getting what you want(employee 2 is gone) but not like what you get(decreased productivity for your other team members.
It’s the result of a low trust society. Yay.
Companies and employees form into these longer term relations in an attempt to reduce the burden of creating these types of task based contracts. Is it more efficient? Maybe, since it is more common than contract work.
> drew a straight line between here's what you need to do to get this much of a raise or this much of a bonus
This is really really hard to do. It's not a good idea (or even a practical idea) to plot out all the work that you want an employee to do in a year, or even in a quarter.
You can prove this to yourself. Try and write down all the things you would like to get done over the next year and how much of a bonus you think you should earn for each thing. If you're in sales this is easy - it's a sliding scale ratio between revenue won and bonus. If you're in product or engineering I suspect you will not find it so simple.
If you've read posts like "Being Glue" (https://noidea.dog/glue), they talk the importance of work that's impactful within the team but very difficult to measure. If companies tried to draw a straight line between visible work and raise/bonus, nobody would ever choose to be glue.
This isn't to say that everyone is happy with how annual reviews work. As engineers we want everything to be black and white. If I pass this unit test, I get a pay rise. As managers who used to be engineers, we'd love to be able to give engineers that clarity, because you're right, they would get hyperfocused on it and we'd get a lot more out of them. But nobody's worked out a way to make it work.
Believe me: if you can invent a technique (and make a product around it) that enables this, you'll make a lot of money selling it to every single software company.
I think it's possible as long as you define the work very generically.
For example, replace "finish task X" with "participate in more code reviews" or "improve documentation". Sure it may water down the review, but at least at the end of the year you can reflect back on your general performance and measure it in some way.
If I tell someone "if you improve documentation, you will get a $10k raise", then I am tying a very well defined reward a very poorly defined task.
"Improve documentation" could mean fixing some typos. Or it could mean documenting every un-documented public class and method. Or it could mean refactoring code to make it easier to use so less documentation is needed.
These all have very different value. In a year's time, I'd have to justify why only one of them actually might warrant the 10k raise. It is *much* worse to tell someone "you were promised a raise... now you aren't getting it", than to never offer or give the raise in the first place.
> As engineers we want everything to be black and white
should not be true of "us engineers". Engineering is balancing trade-offs, and any moderately complex system is a big hairy ball of tradeoffs, some of them (in the best case) unarticulated. I get suspicious of boolean logic applied to real world, especially after working in AI for a long while.
That black and white thinking is characteristic of juniors, and some mathematicians dealing with models disconnected from the real world.
If we could predict exactly what work needed to be done and how long it would take a theoretical employee to do it, this would be great.
But we can’t predict the work with such accuracy, nor can we predict how long it should take to get done. If we planned all of the year’s work in exact detail and underestimated by even 25%, you’d be working 10 hour days every day for a year. It’s not uncommon for projects to take twice as long as estimated, in which case you’d be working 80 hour weeks.
Another problem is that once you set specific numeric goals, people will game the hell out of them. I worked at a company that decided to reward people by the number of bug tickets they closed. Suddenly, the number of bug tickets tripled! People were starting to ship code with known bugs because they knew they could get some easy bug fix ticket closes to raise their numbers. It made outcomes worse even though everyone now knew exactly how to get bigger bonuses.
Finally, there’s more to reviews than just the things that can be measured through metrics. If someone is crushing their tickets and committing a lot of features but they’re insufferable to work with, they shouldn’t be rewarded for that. Good managers should have their finger on the pulse of these softer aspects of performance, even though it will never be a perfect assessment.
It's a complicated problem and it's also more a social/relationship problem than really a technical problem. How people interpret what you've done is much more important than what you've actually don't.
Please don't try to come up with technical solutions for social problems, I see engineers trying to do this and failing all the time.
No one is on the same page about it. The purpose of annual reviews is unclear, misunderstood, and misinterpreted, leading to a broken implementation in practice.
Even when there is an attempt at clarity, emotional kneejerks interfere anyway. ex: Telling someone that they "Meet Expectations" when those expectations are very high just results in them feeling like you've called them "average" somehow, and people outside that process are absolutely going to misinterpret it as "Average" leading to a continuing failure to get somewhere constructive.
Most of the time, there aren't great ways to be usefully objective about performance, anyway, due to missing context.
And having everything be extrinsically motivated is horrible for job satisfaction. It would actually be better to hide the reasons for raise/bonus as much as possible so people want to do their jobs for their own satisfaction (obviously people want to get paid too but it’s best not to have thoughts like “if I fix this bug, what’s the exact $ bonus increase I get”).