Bonuses don't work well because they are too temporally distant from the work you do. For example, typically you get an annual bonus, but that might be for work you did a whole year before (and even more, with the time it takes to decide on bonuses, pay them out, etc.). There's just no way our brains properly can handle reward that is so distant in time. As a result, bonuses end up feeling either "expected" (if you always get roughly the same amount) or capricious/political (if there is a lot of variance).
Add to this that accurately rating employee performance is very difficult, and you get a huge disconnect between effort and reward. What this means it that employees will seek to game the system by identifying activities that give them a high gain in "review score" for the least effort. This leads to rent seeking, politics, stealing credit, and other such activities. Not everyone plays this game, but those that don't are at a huge disadvantage. I've seen people get the same performance rating where one of them did literally 5x the work of the other, just because of other factors.
Bonuses can only work when they are given soon after the work you do and are (at least somewhat) objectively-measurable. For example, sales commissions work well. I think signing bonuses are also pretty effective at luring a job candidate when they have multiple offers, because they get the money right away. But that's sort of a different scenario.
I’ve been working on software for more than a decade now at companies most of which implemented some form of kpi system. Ime nothing you do more than 1-2 months before bonus is evaluated ever matters. You can ship some big things in the beginning of cycle and then hit a rough patch for a month before review and 90%+ of management will hang you out to dry. Or you can fuck around for better part of the year then ship a week before reviews are in and get a bonus/promo
Here's an idea to prevent such perverse incentives:
All employees are part of a pool that evenly distributes a portion of the company's profits.
The employee's incentives become tied to what the company wants most: income.
This might create other incentives like trying to hire fewer people so as to not dilute the pool, but I can't think of a better heuristic than to tie the company's success into each of the worker's success.
Where I work, annual bonuses work partially like that, and partially based on personal performance. It's a target percent of your salary (around 10% depending on level), with multipliers for personal and company performance (ranging from 0-200%). This year my bonus was roughly 1/3 my base salary because the company performed well, and I received a good review. It certainly felt awesome, but at the same time, in a company of 100k+ employees, am I really personally responsible for how the company performs in any meaningful way?
What I'm saying is, I think the GP comment is right, that a bonus scheme is unlikely to materially affect performance.
That type of scheme does however help foster alignment, which can be more valuable than individual output. When the performance of the company is a bigger multiplier than your personal performance, you are financially more interested in the company doing well than your own stardom.
It doesn’t really drive people to work 80 hour weeks, but that isn’t by itself valuable anyway.
This sounds like it'd stop working really fast as the org grows. Low-performing/average employees will start to feel like their contribution to the pool is too small to justify additional effort while top performers will feel resentful that their work gets diluted among people who don't have as much impact.
One issue with this system is that employees are incentivized to increase short-term profits at the expense of long-term growth. (Arguably this is already true of public companies' pressure from stockholders, but for private companies this may make a difference.)
I agree. I wonder how many other people agree with the word "evenly"?
I think it should be proportionate though. If you work for 1 year full-time then you get 1 portion. After all that person is giving approx. 1/45th of their working life to the company. ( Pro-rata reductions for people who work part-time )
It's a pity that the people who choose the remuneration would never go for it. They might not be able to afford their new BMW if they are sharing the bonus pie with the staff who actually help customers.
I agree with the "evenly" bit because it is a bonus on company performance. This is simply saying that employees are invested in the wellbeing of the company, as much as "investors" are.
As an individual, that person should get their wage/salary too, so experience/loyalty etc gets factored into that.
>If the bonus dollars are evenly distributed as proposed then it won't encourage competitive in-fighting that plagues some incentive systems.
I am not sure about that. I think the in-fighting would then come from people upset that other people that "did not contribute nearly as much as I did got the same bonus".
People will always find a way to no be happy about something.
Some people will, sure, but some systems encourage it more than others IMO.
If a company designs their bonus system to encourage individualism and thus explicit competition for scarce dollars, then it creates a game that people think they have to win at the expense of their peers. In these cases, they have a bonus pool (the "budget"), differentiated pay philosophy (pay for performance), and rarely have an objective criteria for assigning rewards (manager decides and makes a case). In that situation you get a lot more in-fighting because it's the way the game is designed.
If you remove the competitive angle by assigning rewards equally (say, at a certain level of seniority you get a fixed bonus %; or, the company assigns a 20% bonus to everyone to share in the profits), then you are no longer giving people a game to win against their peers.
If a janitor gets the same "company bonus" as the rockstar dev - that is ok. The rockstar dev should have a salary/wage to match.
And both are vital parts of a company. You can't have a bunch of co-located people without someone taking care of sanitation.
Of course, you may choose a 100% "remote work" company to avoid that issue but in this instance each employee is their own janitor anyway.
There is a flipside benefit too: the lesser paid people - other devs? You? Me? - will be happy to see the rockstar dev get such a wage/salary because they know they are worth it and it adds to their own company bonus.
And when people are not such a rockstar after all, they will see scrutiny on all sides.
I think they meant that the money would be paid out yearly from profits. Not trickled out to you in "virtual dollars" that may never be worth anything unless your company goes public or gets acquired.
The privately owned org that I started with recently has new comers at a pretty constant 30% annual bonus. Tied to how well the company did but upon conversation with veteran staff, no issues with management paying out or percentage drops in the last 15 years.
They aren't a perfect company by any means, but they seem to be decent from current observations.
Bonus schemes can be extremely effective for retention. Partition comp so that enough turns up monthly that people don't struggle and add a significant bonus every N months and there is never a 'good' time for your employees to leave.
Best I know of was every three months as that's roughly the latency to change jobs - after a bad day in the office, the next reward is still close enough to distract from the alternatives.
Ah yes, "retention". Or: "we know how bad this place is to work, so we have to backload your pay to convince you to stick around once you work that out too".
You're absolutely right. Last company had a sizeable bonus, and while I wanted to quit years prior, 'bonus' and 'yearly 401k match' were always just around the corner.
During my most recent company's negotiation, I was honest I wasn't going to do that again for that reason. Ended up negotiating from a 70/30ish split to a 90/10ish.
Bonuses that can be clawed back will never keep me around. I do generally appreciate my bosses, but I know better than to assume my hard work will be recognized. I work hard because I like being valuable. I stick around for a reliable income.
Also, I generally have low opinions of coworkers who do anything else. Those desperate for recognition usually sacrifice something else in the process.
> Partition comp so that enough turns up monthly that people don't struggle and add a significant bonus every N months and there is never a 'good' time for your employees to leave.
I don't think "ideal" is possible, but if you have to choose whether valuable employees in average stay longer than they'd want or leave earlier than they'd want—longer is better for the business.
I think the same is true of equity compensation schemes.
I understand the theory of aligning incentives, but in practice what I see is a bunch of perverse results. Rest-and-vest as well as all sorts of non-useful stress in times like these.
Equity compensation works for the CEO, where he has a notable and direct influence on the share price (whether that is a GOOD thing is another question).
For normal employees, it's just a form of pay with extra hoops. I suspect much of these started with ways to get "around" cash flow requirements, etc, but now the accounting rules have caught up and it's not really worth it anymore, for anyone.
I’m not sure I agree with you that equity compensation is only beneficial for C suite level people.
I think a lot of people in the Bay Area who work at large and small companies have benefited from rising stock prices and large amounts of their pay being equity. Look at the price of Apple stock over the last few years. Even a normal IC at Apple has made significantly more money due to their equity compensation being granted over 4 years.
There are also many people who have been normal IC level employees at companies like Snowflake who gained tremendous wealth by them going public and being paid in equity. Sure there are some losers, but I think over a large sample size employees generally win by being paid in equity vs the cash price at the time equivalent.
> Even a normal IC at Apple has made significantly more money due to their equity compensation being granted over 4 years.
The equity compensation at Apple is ridiculous. For one they're super stingy with it. Thanks to the infestation of stack ranking only the top quintile or quartile of employees on any team (by whatever performance metric the manager decides upon) will get any RSUs. Upper management has no problem if all the RSUs allotted for a team go to only a single team member. There's a reason Apple really hates the idea of employees talking openly about compensation amongst themselves.
This is just hindsight bias. If these employees had been paid the equivalent additional amount in cash would you have suggested to them to invest all of it in their employer's stock?
No: the point of hindsight bias is that you’re able to make better decisions when you know how the future will go. Hindsight bias that those companies will all grow in that five year period doesn’t change the way you would decide if you should invest in your employers stock today.
You can approximate that with being paid more and buying company stock - but most would recommend NOT doing that. Pre-IPO stock options and grants are a separate discussion; I was thinking mainly of the equity portion of Google, Facebook, et al.
It ALSO allows them to "pay" people the same even though total comp may be very different, (this looks good on reports about pay equality) AND it allows them to institute pay cuts without ever having official pay cuts.
Having a portion of pay that automatically and transparently scales with company performance seems valuable and sensible. When the company outperforms expectations, the gains are shared. When it underperforms, the pain is shared.
The problem is that each individual employee generally has vanishingly small individual impact on the company's performance. That can really hurt morale when the stock price goes down, often even when the company is doing just fine, or even well.
IME, it doesn’t hurt morale worse than the company instituting broad-based salary cuts would, which is an alternative structure to accomplish a similar shared-success model.
Sure, but we're not talking about salary cuts or a shared-success model as the only possible alternative. Why not just do away with equity compensation or bonuses entirely, and pay people more?
(I mean, I know why; equity compensation is a much cheaper way for a company to compensate employees than extra salary or even bonuses.)
I think shared-success is a fair model (regardless of whether or not it's good for morale or productivity) when the company is small enough where you can at least somewhat understand how an individual employee's work contributes directly to revenue. But companies grow past that point, sometimes quickly.
Having said that, I think I made out very well with equity comp at my last company, much much better than I would have done if there was no equity comp, and we all just had higher salaries. But I think that outcome isn't common, especially when you only consider employees in non-executive positions.
Share price is driven more by future expectations than by past performance. Your record profits are nice, but are history at this point.
What matters is expectations of profits going decades into the future and the relationship to the risk-free rate of return (which has risen dramatically recently, due to the expected future devaluation of the dollar).
That pretty well reinforces my point: share prices do not rise because employees have done a good job, profits do. RSUs are not, therefore a shared success model.
And explicit profit-sharing exists and is found in some companies; but rarely Silicon Valley ones. (They seem more common in factory/service companies)
> Another explanation is that the reward makes the work seem distasteful. “If they have to bribe me to do it,” a person might figure, “it must be something I don’t want to do.”
Logically this also applies to any form of compensation, say, your wages. (Hence the derogatory term "wages slave" from these who can afford not to depend on wages.)
The old recipe for that is to "pay your engineers so much money they won't know what to do with it" (can't find that quote from 1980s), but the realty market has adapted and likely made this a largely inefficient strategy.
I think fixed wages have less of an effect on this because you get used to them so you don't really think about them. Especially when you aren't working hourly, the relationship between the work you do and your pay is often very tenuous. So it can feel like you are doing the work for its own sake.
For me, getting paid per hour is anathema to my enjoyment of the work. It’s a perverse incentive if I get things done faster than expected and a source of worry if things take longer. I’d much rather get paid per project, or get paid salary with TOIL for overtime worked.
Second, they make a lot of claims that at least go against my understanding of psychology - and with no citations, only mentioning their own book as a "source" at the bottom.
All the discussion about bonuses being counterproductive rewards for performance proceeds from the assumption that bonuses are intended as a reward for performance.
Actually, bonuses are mostly a way for a company to be flexible in the amount paid to employees. If the company has a bad year, cutting salary will have people looking for the exits... but they might accept a smaller than usual bonus.
I once received an employment offer from a FAANG that was weighted heavily on the annual bonus. Like 1/3 of the total comp was 1st and 2nd year hiring bonus. I don’t get how that’s appealing since it seems like a guaranteed pay cut once those are done.
Was this Amazon? Generally equity would replace the cash bonus in subsequent years. Compensation then becomes somewhat uncertain, but if it became a huge pay cut at that company and nowhere else, it seems expected that lots of employees would quit. That might be partially intentional.
At Amazon, the initially granted tshares vest in a backloaded way. For every 6 months after you start, the vesting is (0%, 5%, 0%, 15%, 20%, 20%, 20%, 20%).
The large signing bonus for year 1 and year 2 is structured to make up for the back-loaded vesting.
I was thinking after those 2 years were up I could go interview again and get a competing offer and then ask for a retention bonus, but I don't know if they'd do that and it seems like a lot of work.
One problem with boni is that they become mali if one does not get what one expects. This is especially problematic when it affects the better part of a company's workforce. Then the company is demotivating their best employees.
A related hypothesis: There is an asymmetry because overvaluations do not have an equivalent motivating effect as undervaluations have a demotivating effect. Boni without an objective measure are therefore on average rather demotivating.
Within the context of software development, for best results get rid of the managerial barrier between engineers and the business; the original intent of Agile.
Except “agile” has evolved to adding 3-4 layers between the engineers and the business because the engineers tend to build products that are great for them, but lousy for the business. Whatever the intent was, the effect sucks.
I am sure I will get downvoted here and don’t care. Three decades of experience watching unfettered software engineers get it completely wrong overrides your derision.
Not at all, the reason businesses feel compelled to add that buffer between software engineers and the business is because software engineers tend to make lousy decisions in terms of understanding the actual product requirements…at least in my experience.
Ahh, your assertion is that things have improved with the extra layers. I’ve been developing software professionally for 27 years. I respectfully disagree with your assertion. I’ve seen a steady decline in the quality and efficiency of software development since these extra layers have been introduced.
I’m guess what I am saying is that businesses feel they need the extra layers because software engineers today are generally terrible at translating the businesses requirements into a viable product.
But, I think it speaks to the general quality of the software engineers produced today more than in the past. 20+ years ago, I think software engineers were more talented and perhaps more capable of directly interfacing to the business without the middle men.
So it’s kind of what came first, the chicken or the egg.
In the 80s and 90s, I recall 9/10 pro devs were exceptional. Today, by my generous estimate, I’d say 2/10 would measure up to those folks I worked with back then.
Again, thats just my experience, but I am an old man nearing the end of my career who yells at those damn kids who keep playing on my lawn nowadays, so take it with a grain of salt.
Bonuses (and the like) seem very different than piece work, which the author conflates. I worked for a few years doing piecework and the direct correlation between my productivity and my income was perfect as far as I’m concerned. I became better in every way - faster, better quality, more sensitive to the owner outcomes (profit).
In my subsequent 25+ years working in software (QA, Developer, Architect, Executive) the closest thing I’ve found is ownership participation (Options, ESSOP, etc.) which is too often a bad deal for employees because of various investor hijinks but generally helpful in creating alignment in both growth and efficiency. I’d prefer a bonus calculated based on the sum of growth and margin since it likewise creates alignment on the things that matter (in capitalism) but it’s very rare to find others that agree and support such a program at the investor/owner level.
"The free market" is not synonymous with "money as the primary motivator". All that a free market means is that we let natural interactions between people and groups of people drive the economy. It doesn't rely on any particular framework for how people choose what they want or how behavior is actually motivated.
As an example: I choose to work at a job that pays less than I could get elsewhere because there are intangible benefits that matter more to me. I'm able to do that because the market is free. Within the company, I fight to keep those intangible benefits prioritized because I know that they are a major selling point that will get us engineers when we can't afford to pay SV wages. That's the free market at work, but money and incentive schemes don't factor into it at all.
You or they misunderstand. The fundamental premise of the free market is that, in a large and complex system, attempts to exercise central control tend to go badly relative to just letting people figure out what they want and pursue that. It's not money specifically that is the incentive to work, it's self-interest in general. This is perfectly compatible with Alfie Kohn's idea that intrinsic motivation beats extrinsic motivation—intrinsic motivation is a form of self-interest.
The primary ideological conflict between communism and the free market isn't over money, it's over whether it is possible for a centrally-organized economy to outperform one in which people are left to pursue their own self-interest as they see fit. The article doesn't address this question at all.
(As an aside, central control versus free market is obviously not an either/or choice, and most countries' policies end up somewhere in the middle.)
This article argues against a particular form of compensation, not compensation in general.
And in any case, "the free market" != "people are only motivated by money". All "free market" means is that voluntary exchanges are not regulated or restricted by the government. Indeed it has nothing inherently to do with money, if you want to trade your labor for a pile of corn that would be a perfectly fine free market transaction.
I just rephrased the findings in the article at a macro-economic level and came to the conclusion that the free market supporters were not telling the whole truth. That's all.
Add to this that accurately rating employee performance is very difficult, and you get a huge disconnect between effort and reward. What this means it that employees will seek to game the system by identifying activities that give them a high gain in "review score" for the least effort. This leads to rent seeking, politics, stealing credit, and other such activities. Not everyone plays this game, but those that don't are at a huge disadvantage. I've seen people get the same performance rating where one of them did literally 5x the work of the other, just because of other factors.
Bonuses can only work when they are given soon after the work you do and are (at least somewhat) objectively-measurable. For example, sales commissions work well. I think signing bonuses are also pretty effective at luring a job candidate when they have multiple offers, because they get the money right away. But that's sort of a different scenario.