You should almost always avoid working for low-margin businesses. Sure, if they are new and growing, there's an exception, but if their whole growth plan is premised around being low-margin, it's not going to be a great place to work and advance your career. Expenses such as compensation are always on management's radar.
I actually didn't get a job at Amazon because my pretty normal base salary requirements were higher than most of the execs in the company.
The deer in headlights wide-eyed look on my recruiter's face (after I had finished all the interviews) when I told him what I currently made in a smaller metro and what I was looking for is something that will be seared into my memory forever.
I can't remember if the word "salary cap" was used or not, but it was definitely implied.
I offered to accept a lower base with alternative compensation options (even vacation days), but it was already too rich for their blood.
I think they gave it some serious thought, but I got a call a week later that they simply couldn't do it.
I'd describe AWS pay as "competitive". Most all gigs provide you a decent signing bonus and stock grants. This lowers your base salary as all these items together are taken into your total compensation, which I believe is pretty much the norm up in Seattle for AWS/Google/Microsoft.
Your signing bonus is pro-rated and you pay it back if you quit before one year, and your moving expenses are also paid back over two years if you quit before then. Your stock grants vest heavily in the back-end over four years. Every year you get a review and new grants are included in your total compensation package as your "salary". It's a sort of lock-in system that helps keep you invested in the company as you keep earning stock etc, but that was my perspective. Nothing exorbitant, and that fits with the Amazon ethos of "frugal".
Yeah, that jives with what I was told. I think the stock grants mature over 4 years? I think the schedule was kind of interesting as well but I don't remember what it was, it wasn't 25/25/25/25 either.
To be honest, I think it was a miss-match in positional terminology. I was going for a PM role, which in my current career path is basically the CEO of a specific project or program for part of a company. (These programs can be very big, the largest program I ever worked on as part of the PM staff was north of $130m/yr. but for a wide variety of reasons I prefer the smaller $5-20m work, much more fun to run). Typical compensation is a base salary + specific performance bonuses + contract win bonuses (+ large option grants as bonuses in the kinds of companies that do that sort of thing).
I got the sense that at Amazon PM roles are much lower down the totem pole positionally and act more like coordinators in the organization and don't have as much agency as I was assuming. I was honestly just looking for a change of pace and probably would have accepted whatever they offered even if it was quite a bit lower than what I considered my market rate, but I fouled it up early on and didn't recover well.
If they're like PM roles in the rest of the Internet industry, then yes, they are basically CEOs of a specific project or program for part of a company. The problem is that in most large software companies, projects are generally quite small. So a PM might be in charge of an engineering team of 4-6 engineers that are working on one tiny widget on the page. The product as a whole is usually the responsibility of a VP.
Anecdotally, Amazon's pay scale is only competitive with other companies in the area if you ignore on-call and only look at starting salary/bonus (including stock compensation).
In silicon valley, they just pay a silicon valley wage, which belies the fact that they are in the south bay. Consequently, managers here have constant trouble convincing San Franciscan talent to work there - they just aren't very competitive. I have not seen anything spectacular in terms of a signing bonus, either.
Furthermore, they don't increase your salary over time, so after a while, that OK starting salary drops below the average.
That doesn't refute the original argument. Amazon might very well be a low-margin business that pays competitively (they gotta attract that talent) while working at a "grinding pace" (as in the article).
Ex-amazonian here. Another problem is if your manager doesn't want you to get promoted. For whatever reason, sometimes people don't click, and I've seen the careers of some extremely talented developers be hampered by a bad or non-relationship with their manager. It didn't matter that their entire team, and other teams viewed these people very highly, and sought out their advice and consultation on a number of matters - they still couldn't get promoted unless their manager signs off on it.
I worked for a fairly large company that previously required your team manager to approve of you moving to another team. You could spin wheels all you wanted, but if they didn't like you for whatever reason they could immediately stop any lateral movement. This wound up pushing a lot of employees out, they'd eventually quit because they were miserable. Couldn't get shifts they wanted, bootcamps, training, unfairly poor reviews, etc.
It was one of those "who.. thought of this?" ideas. I believe they recently removed it.
Variations on this are pretty common, although what I've seen more is that you're obligated to inform your manager before seeking a position on another team- your manager has no power to prevent it from happening.
I don't know why a manager would want to prevent this from happening though. If I were a manager, and one of my employees said they wanted to switch teams, the last thing I'd want to do is force them to stay on my team knowing that they didn't want to be there, even if I was feeling vindictive towards that employee.
I've heard that the policy requiring informing the manager was motivated by a desire to allow employees to switch teams if they wanted to, but at the same time to dissuade managers from actively poaching employees from other teams in the same company. Also, it gives the manager a chance to possibly make things right with the disgruntled employee, instead of just losing that employee with no warning.
It's not a perfect system at all, but I can at least understand the rationale. It totally doesn't make sense that a manager would be able to prevent it from happening though.
I feel this same thing happens at almost all companies I've worked for. I had to leave one of my favorite jobs at a company doing really well because I was never going to get to work on what I wanted, in the group I wanted...
It's amazing to me how much money companies can waste hiring new people - albeit a not unwelcome redistribution of profit from shareholders to developers - yet once you're hired you're stuck in that role unless your manager signs off irregardless of other internal opportunities.
I wonder if having a more liberalized internal labour market wouldn't be beneficial; it's destructive in so far that being undermined by more popular projects keeps siphoning off your best talent, but if the alternative is "quit and find another job" it strikes me that keeping that person within the organization may be more beneficial overall.
It's not so much manager doesn't want to promote you as manager has no role to promote you into. Sometimes you can create a role and demonstrate to your manager that you are willing and responsible to manage that topic, and perhaps it needs added investment to grow that topic (more money and more people) which you will manage ... in other words don't wait for your manager to create the role to promote you into. Most managers are happy to team leads with this kind of initiative and will support them.
If you create the role but still your manager is unsupportive about it then you're probably in a team with a very narrow scope and there's just no further room to grow in that team, so time to move on.
I've heard that at Google, the team votes for the person they would like to see promoted. This seems to be a pretty good idea from the outside - can someone comment on this? There must be some Googlers lurking...
If you want to seek a promotion, you can nominate yourself or others can nominate you. When that happens, a packet is assembled for you. It includes your self-review, and reviews from a number of your peers as well as your managers.
That packet goes to a committee likely made up of people who don't know you who then reach a decision about your promotion.
That's a simplification, but that's the basic idea. Your manager has a large influence on your success: to get promoted you need to show accomplishments, and that depends on you being given important work to do. But, compared to other companies, the process is much less biased by your manager. Your peers have a huge influence on your ability to get promoted.
So if I was a Googler, I'd apply for a promotion just to see what my manager and my peers thought about me - both pros and cons - and use that to improve my performance.
All the time. Conventional wisdom at Google is that you should always apply for a promotion, because the cost to you was zero and the benefit was potentially very large.
...that said, I didn't apply last cycle, because the form was changed so that promo candidates have to fill out a more in-depth self-assessment and get more in-depth peer feedback, and me and my peers were heads-down on a deadline at the time. I suppose one way to deal with the externalities of everybody always going up for promo is to make the process easier on people who don't.
BTW, there are regular performance reviews at Google - you can (and should) get your peers' feedback without actually going up for promotion. I did every cycle until I made senior, then I stopped caring quite so much. Promotions beyond senior have a large "impact and leadership" component to them, and so you basically know what you have to do to get promoted, you have to lead a major project that succeeds.
"Ambitious employees tend to spend months having lunch and coffee with their boss’s peers to ensure a positive outcome once the topic of their proposed promotion is raised in an OLR."
Sounds like all those Microsoft stories that came out a few months ago.
its a trend others and I have seen elsewhere. Far too much a popularity contest, trying to ensure during your daily and weekly routine you don't piss off someone who might be sitting on the review panel.
It has taken getting good work done and placed it second to making sure to not tread on any toes, even if that means things that should be done don't get done.
Article rings true from my perspective as a "level 5" manager in Amazon's fulfillment operation.
Remember that Amazon core value to "have backbone" and "don't be afraid to disagree"? This is where that would bite you in the ass. That Senior Operations Manager you disagreed with that one time in a staff meeting in front of the GM? Yeah, they just sunk your promotion during the OLR because your direct manager didn't want to put their spine out there to get snapped.
Not sure how different it was in Seattle, but in order to get that promotion, you had to tread lightly and make sure everyone thought of you as the perfect little angel.
Maybe the problem is promotion altogether? Work your ass off for 1 year and then, you may or may not win the promotion lottery. I guess a more flexible style of salary increase could be created?
Work on a project for 3 months. If it attain its goal, everyone on the team receives 0.20% increase. This way you want to make sure your team will succeed.
I'm just thinking outloud here. It could be a rubbish idea.
Much more simple than that. Just take a cut from the profits your project generates or supervises. If the project doesn't generate or supervise anything just get a cut from the overall budget which is allocated based on how valuable the project is to the company.
Favors near term profits far more than long term growth, which you could argue is not good for the longevity of the company. Plus, there would be competition over the seemingly most profitable projects. And there are projects that don't generate any profit, but has benefits that are very difficult to quantify.
But if that's true then it applies to shareholders and execs as well. Should we change their compensation plans? Yes there will be competition over the profitable projects and that's how it should be. Projects without profit don't exist. If the project doesn't have profit then the company wouldn't allocate budget. If they allocate budget then the project is profitable. It is up to the project leaders to have the fight and justify budgets if it's not that obvious.
Depends on your industry. In finance there are many, many projects which have to be done for legal compliance. There is no 'profit' for the company; the goal is to be permitted to remain in business.
Funds have to be allocated to do the project, but it brings in no new revenue. It might actually incur new costs, such as ongoing storage for audit logs.
Well, any prevented loss can also be interpreted as profit. So depending on what kind of money is at stake you can reasonably claim a cut. You provide insurance and your cut is the price of the insurance policy. It's a well established business.
Not arguing at all that things have inherent value, but the challenge is having an objective way of putting a dollar value to those activities.
For example, does a backup generator add any value if the electricity never goes out? Do you only assign it 'profit' if electricity goes out? How do you calculate the value of the downtime if electricity does go out? Does it encourage people who've worked on setting up a backup generator to 'wish' for outages?
It's a challenge. Even if you do have answers to all of these questions, there will always be levels of subjectivity resulting in politics, and the effort involved will be itself an administrative workload of which the 'profit' would be difficult to assess.
Then there would be no technical support staff - upgrading computers generates $0 new revenues and has significant costs.
Clearly there is value in not forcing developers to use 15 year old machines, but it isn't so clear to quantify.
Since whomever 'quantifies' this data then has enormous influence as to what projects are worked on, you risk just shuffling around the politics: instead of fighting over what projects to work on it becomes about how many dollars each project is worth.
Unless you discount the expert time saved per year. This is actually a real metric. How do you think they calculate the budget for tech support? They measure time wasted in tech issues, multiply by the expert wage and then start doing a binary search up and down to see what optimizes the cost at average salary increments. Of course politics corrupt the process eventually.
I worked at EA when they tried a similar model. It was astonishingly effective at slaughtering any technology shared across game teams. Large companies get economies of scale by being able to reuse technology across multiple revenue generating products. If your compensation system can't accommodate that, you'll lose it.
Well, it depends on the length of the project. 0.2% for 1 month work equals to a 2.4% raise by the end of the year (12*0.2).
The thought would be like this: You give impacting projects to your best talent. Higher the impact, higher the % after the completion. So your less efficient talent would have a salary that follows inflation while people thriving could get very nice salary increase.
If you want to step up, you request bigger responsibilities. Management test you out by giving you 1. You get better % on it. If you fail, you go back to step-1. If not, you just open the door for bigger salary increase and more interesting projects.
I think the most interesting part of the article is when it mentions that anecdotes about an employee's performance most-often dominate these OLRs. It seems that a company with as concrete of a business goal as Amazon's would have the ability to use a more objective means of measuring employee performance.
The nice thing about knowing everything going on in your company is being able to judge it easily when it exceeds expectations. The shitty thing about standardized metrics is gaming those metrics.
Non-Amazonian here but I can tell you this so-called OLR process is same as how most companies with more than 100 employees deal with the promotions process.
In most companies your line manager does not solely decide your performance rating and promote you based on that, but rather the manager should be telling you how your are performing, how to improve and grow, and once you have proven you can manage added responsibilities then your manager suggests and advocates for your promotion. The promotion candidates are then considered across all teams during an annual or semi-annual review process.
This practice ensures that a lenient manager isn't promoting their junior interns to VP roles, every team has the same performance standards, and managers of other teams can endorse or disapprove of someone being promoted based on whether they work well with other teams rather than just performing well within their own team.
I think that there's a lot of subtlety that gets lost talking about the corporate structure in Amazon. The levels were more or less correct, so far as it goes, but it's not the full story. For instance, a senior developer could be level 6, while his manager might be level 5. That doesn't mean that the dev gets to override what his manager says; they're different promotion paths.
This is key. If you want to work as a dev at Amazon, you'll go from level 4 to level 7 as a dev. I don't know if devs are ever integrated into the corpocracy, vice presidents and so on. I suspect that they generally aren't; at least, I've never heard of it happening.
A manager's levels are different. I'm not and have never been a manager at amazon, but the responsibilities at each level are different from the dev side.
Any organization is going to have a very limited number of top slots, and so the larger it is the harder it will be to climb into one of those positions.
In many organizations a promotion will depend on your supervisor's ability to sell it to his superiors. The system at Amazon just makes it more noticeable.
I'm highly opposed to establishing quotas, either for promotion, demotion, hiring, or termination. In all cases, quotas establish a destructive environment.
Whenever I read one of these articles, I wonder which medium to large size companies actually are good places to work. This is from a "startup type person perspective" -- i.e. largely defined by work or career, even if you don't want the financial risk of a startup. Lots of places are fine if it's just a job.
Apple seems like it could be awesome if you're on the right time, and horrible otherwise, which is about the best I've seen of anywhere.
By definition, any job where you're not revenue generating and in the main (and especially main and growing) line of business also sucks. I guess there are limited exceptions when it is clear a new product could become the main line of business, and you're early in it.
Amazon, Microsoft, Intel seem painful at best, and Google seems like it had a rough few years with a mass culling of less important products.
On the other hand, even the most pathological organizations (government!) can be ok as a consultant, if you're on an important enough assignment and have buy-in at the right levels. If nothing else, it's an amazing opportunity for comparative anthropology.
Because promotions are generally to a new role, and have been perverted at some places to simply be a new title. Want to get promoted from a Developer to a Developer-2? Those are unlimited.
Want to get promoted from a developer-15 to Amazon's CTO? Those are limited.
The Developer to Developer-2 (in Amazon parlance, SDE-I to SDE-II) is VERY limited at Amazon as well. OLR doesn't just hold back managers, it holds back everyone, and it badly hurts their competitiveness in the labor market. After 18 months as an SDE-I there, I received a 4% raise, which was huge, for Amazon (though cost of living in my area went up substantially more during that time). But there was never a chance I was going to be promoted at that time, even though I had been very close to being hired as a II, and if I'd been interviewing then instead of being reviewed, I would have had the II title and compensation. So I left a couple months later (mostly for other reasons, but I won't pretend this wasn't a factor), and got that at my new job. Their loss.
Skip 9 but you have 11? Obligatory "this ladder goes to 11." Really, what's the point of removing a wrung on the ladder if you're only going to put it back at a different/higher place. You could have steps 2, 3, 5, 7, 11 and then say you've removed 5 different wrungs but have you really?
Worked at a firm that used the same method for raises and promotions. It builds animosity and decreases synergy between teams. When there is a limited pool of reward money and you know the split hinges on the perception of your department there will be a lot of blame games and outright sabotage.
How can it be harder to climb Amazon's corporate ladder unless they have less management or do more outside hiring at a senior level than other companies? Somebody has to get promoted to fill the positions, right?
Why don't companies just ask employees what they would like to be doing, and pay them to do that? It's stupid to promote someone into a job they don't want, or keep them doing a job they don't want.
This type of review system also exist in my job and I think this is crazy, as they expect you to please every other manager in the team, which further leads to favoritism and bootlicking.
That's not really accurate. 4 are usually college hires, at 5, other job families can split off. Engineers, Managers, TPMs, Product managers, etc. they're all in parallel. So you have Level 4,5,6,7 and higher engineers as well.
Maybe it should be difficult? There's usually nothing
more useless than someone who wants to beome a Manager.
What even surprises me more is the amount of people out
there who really don't care about the money; they want
authority. They want a Title in life. They need a title
in life?
I have noticed the one factor that can produce a good
Manager. That one factor is birth order, and it's the
first born. If you're an only child, sorry you are one
of the worst. But the first born has been conditioned
into not abusing power, and just getting the job done.
I used to get angry when I read about supposedly progressive tech companies (Amazon, Microsoft, et al) having psychotic, brutal review processes, stack ranking, internal mobility issues, closed allocation.
Now, I'm more inclined to laugh at their rookie mistakes. These MBA/McKinsey hotshots have been (very expensively, I might add) setting up the same exact defective corporate culture (in a variety of different industries) for 25 years, and I (like quite a few on HN) could single-handedly do a better job just because I, unlike the McK morons who think stank-ranking's a good idea, understand talent and what it takes to succeed.
Any company that relies on closed allocation and rank-and-yank is fucked in the long run.
This article is poorly sourced and fails to establish any evidence for it's main point, that it's "difficult to climb Amazon's corporate ladder," which is probably true considering it's a Fortune 100.
Instead, the article paints a picture that Amazon as a place ruled by fiefdoms and popularity contests, but again without any actual evidence, no you have to buy the book for that.
I have a feeling that if I read the book, that I would find that there is nothing here salacious, but that's not how you sell a book.
"In dozens of interviews ranging over two years for my book, The Everything Store: Jeff Bezos and the Age of Amazon, employees often sounded... But in my interviews with rank and file employees... No one I talked to..."
Just because Wikipedia has a "no original research" policy doesn't make first party research "poorly sourced".
Sounds like if one wants the quotes, buy the book. This article seems to be a synopsis of the research, so one wouldn't expect all actual quotes.
// Wikipedia is relevant because their "Verifiability" page is the top result for the phrase "poorly sourced" and Wikipedia tends to be the only context I hear the phrase. Outside of Wikipedia, "poorly sourced" doesn't mean how extensively phrases are in quote marks. it means where the information came from.
In that case, it's not an article, it's an advertisement.
the term "sources" and "poorly sourced" are also often paired with "source confidentiality", "source checking" and other similar terms are often subjects in press and media establishments of notable quality.
I imagine Wikipedia has co-opted this kind of standard operational diligence in an effort to not suck. Unfortunately, in my opinion, this article does not burden itself with such impedimenta.
