Hacker News new | past | comments | ask | show | jobs | submit login
Leak of Microsoft Salaries Shows Fight for Higher Compensation (onezero.medium.com)
206 points by milligan on Sept 14, 2019 | hide | past | favorite | 137 comments

It seems like a strange premise to share salaries on the basis that if we both have the same experience and work at the same paygrade, we are worth the same amount of money - or even for me to know the range so I know whether to ask for a raise.

The underlying problem is still that it is very hard to accurately measure someone's worth as a developer (or manager for that matter) and that there are many who are "experienced" who have not improved in any meaningful way over time and that's why they might not get a rise or a promotion. Pay grades are quite arbitrary and personally, I don't think that a lot of these "super coders" are worth as much as they think unless they also do management (except perhaps for really niche skillsets).

At the engineering departments of most large tech companies including Microsoft, people at the same pay grade are supposed to be making about the same amount of money. The idea is that if you deserve a raise, you deserve a promotion too. So your job title and your salary aren’t like two different things.

Differences arise for many reasons. The most common in my experience is people who negotiate a higher starting salary. You don’t want to just give them more money forever so every year they get their salary tweaked back a bit toward the average for their level. There are also some bonuses that operate outside the normal system.

> At the engineering departments of most large tech companies including Microsoft, people at the same pay grade are supposed to be making about the same amount of money. The idea is that if you deserve a raise, you deserve a promotion too. So your job title and your salary aren’t like two different things.

This is not true. At FAANG companies and Microsoft, the total compensation bands for levels at or above "Senior SWE", "Senior SDE" are really wide. This is by design, and its widely communicated that this is the case. Most of the variation is in equity grants and bonus, rather than salary - if your management chain recognizes that you made a huge impact last year, you'll probably get huge equity refreshes and a giant bonus, although your salary might not rise by much.

Promotions are a separate thing, because promotion beyond "Senior" implies a change to your role and expectations (more leadership, more "people coordination" even if you're not a people manager, etc.) It's totally possible that you're making an amazing impact in your current role in a way that merits additional compensation, but you're not meeting all the criteria for the next step on the ladder. That's fine - you're still going to get paid a huge amount more than the average for your level.

> This is not true. At FAANG companies and Microsoft, the total compensation bands for levels at or above "Senior SWE", "Senior SDE" are really wide.

Ehh, not really. Up to the director level (well above senior), if you give me someone's level and last two performance evals, I can probably tell you their TC to within 5%. There's certainly a wide variance within senior and above levels, but it's not as bonus based as you suggest, and sustained high performance will often (although not always) lead to promotions.

It's not like there are two senior SWEs with one making 2x what the other does. At least not at Amazon, M, or G. Since netflix doesn't have levels, that's possible, and my understanding is facebook has huge bonuses on occasion, but I'm also pretty sure you'll just end up getting promoted if you get too many of those.

Well, yeah - if you control for performance ratings, then pay is normally similar. The pay bands are wide due to pay for performance. Within a given level, employees with really high performance will be paid much more than those with average performance. At the FAANG company I'm familiar with, there are definitely "Senior SDEs" making ~double the average "Senior SDE" comp, because they consistently get really high performance ratings.

helllll no this is 100% not true. I worked at Google, and there is an internal spreadsheet with compensation that is contributed to on a voluntary basis, and has n=1000+ for many levels. The variance is far greater than 5%.

*for Googlers, you can find this at go/salarycomp or something like that, I can't remember

I'm well aware of it. It doesn't include performance ratings. Which I mentioned ;)

well, if it's the same company and you think your work/participation is of similar worth you know then whether they pay a fair price for your work. and if you're a grad you know, what your degree is worth somewhere else.

and generally that's what this is about: creating a job real job market, where both parties can have knowledge of the environment beyond a "you are not good at salary negotiation"-point.

probably we should just have laws that every interview invitation has to include the salary they are offering (better yet: write it into the ad)

California actually has this. Employers must disclose to interviewees the salary range for a position [1]. The big loophole in this is that for tech companies a significant portion of compensation comes from equity (ex. RSU, options, etc). The law is relatively recent and I'm hopeful it'll improve over next few years.

[1] https://www.forbes.com/sites/meghanbiro/2018/09/13/californi...

This is a big lie. It’s not hard. There are no 10x developers. Most people doing a job of the same title in a similar team are likely worth the same money.

People on HN say all kinds of crazy nonsense about it being hard to measure software engineer productivity. It’s nonsense that helps suppress wages and keeps the situation set up as a fight between engineers who should be cooperating for mutual betterment.

Companies don’t approach measuring the worth of any employee with any sort of fine-grained metrics except in highly commoditized jobs.

Also, as a software engineer your value is not all (or even mostly) from writing software. There’s reviewing resumes, attending recruiting events, evaluating new technology the company might adopt, filling out forms, sitting in meetings, sitting in diversity training, etc., etc.

You might even be adding value just by being a name on a sheet of paper. Your headcount tells a small portion of a story that wins the next funding round or puts the company in a better position with an acquirer or gets the company a high rating from a regulating agency or brings hope that causes other key employees to stay at the company or whatever.

These are reasons why some people make 100% bonuses and millions in equity while others get 15% of a low-ball base salary in options they end never able to exercise.

The reason pay is wildly unequal is not because it’s hard to measure productivity. Companies aren’t doing that anyway. Even if they were, they’ll do anything to avoid paying what you’re truly worth.

Pay is unequal because developers don’t know their worth or what forms of compensation are negotiable, because it’s all kept very secretive. Developers don’t develop good negotiation skills and they let themselves often very foolishly believe less tangible and less enforceable properties of a job (like “interesting”) are worth a big pay trade-off.

It all boils down to the same thing. We have crappy pay (compared to the real value we add) and we get stuck in open-plan hell hole offices and on and on because developers won’t say “no” to these jobs and seriously negotiate under the willingness to walk away if conditions and pay don’t improve.

More transparency does lead to improvements for every one and also exposes discrepancies that female and disabled workers face.

Ironically it also leads to an increase in C level compensations.

> discrepancies that female

That is not a given, you have to show proof that there exists a discrepancy in the first place.

I though we where supposed to answer in good faith on here, however:

I personally know of surveys that have proven this both by gender and disability.

Did they take into account the fact that women usually work in jobs that pay less compared to men?

Yes this was a tech company so like for like comparison

Could you cite those studies?

Levels / Pay grades are quite well defined at most companies. Nearly every company maintains a career leveling guide with specific criteria that should be looked for at a level. The difficulty is in evaluation of that criteria. People get mis-leveled all the time. Additionally there's pay grades can be very wide and most levels have significant overlap. Why? Although someone may have been brought in at level X based on interview, they may be performing at X-1 due to other circumstances (personal situation, rest-and-vest, etc). This goes the other way too. Most companies will promote quickly if a candidate was under-leveled coming in and is performing at the next level.

Fully agreed that the notion that pay is (or should be) a deterministic function of a couple of visible variables is ludicrous. But knowing the range can be useful, since you can also have a qualitative sense of how productive you are and how much weight you pull relative to most others at your level.

Obciously you might be over- (or under!) estimating your ability, but that's something you need to resolve regardless, and a rough sense of where you fit in the salary distribution is independently useful.

If you're at a big company and want a significant compensation increase outside of what you'd get from the normal performance review process, the only way to do it is to interview elsewhere and get a better offer.

If you succeed in getting a better offer, you can take it to your management chain and ask them to match it. In my colleagues' experience, they frequently will match it, but not always. In particular, if there are certain firms that have a reputation for "overpaying" for engineers (e.g. Uber in 2016-2017), there might be a blanket policy of "we don't match offers from firm X".

If your offer gets matched, all is well - enjoy the money, but don't expect big raises through the normal review process for a while, since you're probably at or above the expected pay band for your level.

If it doesn't get matched, you're in a tough spot, because your manager may view you as a "lame duck" who is interviewing elsewhere and liable to leave. If you choose to decline the offer and stay and your company, then handling this perception will be very important.

A side benefit to interviewing around and receiving offers - you may realize that your aren't as underpaid as you may have thought. I've had a few colleagues who were unhappy about their pay and chose to interview around, but were surprised to find that the only higher offers they could get were at firms with a much more difficult work/life balance. They realized that their current arrangement was actually pretty good!

Anyway, all this is to say that exercises like this don't have any effect unless people are actually willing to walk. Microsoft already pays for compensation survey data from across the industry - they know how much people are paid, and have set salary bands where they think they're competitive. Unless you're willing to take another offer somewhere else, complaining about comp just hurts morale on your team and makes people less happy.

If you get another offer, bring it to your employer and they don't match it or at the very least come up with a counter you both agree to, then you need to quit and take the other offer. That's the risk of this move. As you say, you will be viewed as a lame duck at the very least and possibly even get fired. Even if your company agrees to match, they may see you that way and have plans for you to leave as soon as you transfer all your knowledge. This is a power play and it can backfire so it's only advisable to do this if you're ready to accept the other offer and move.

I lived in Finland of a while. Tax declarations are public information there and you can find out the salary of anyone pretty easily. IMHO that kind of transparency is actually a good thing to have and helps prevent unfair practices where e.g. men get more than women or nepotism gets you access to better salaries.

We have the same policies in Norway, with open tax registers. It was made for transparency.

But it's a bit flawed, as you can't see things like tax deductions, so the number you see is just a ballpark figure, as much information is hidden.

And besides, a lot of entrepreneurs, rich people, etc. will almost guaranteed have a holding company, and get dividends at lower tax rates.

>Tax declarations are public information there and you can find out the salary of anyone pretty easily.

This true of almost all of the Nordics. For example, you can perform the same request from Skatteverket in Sweden.

>IMHO that kind of transparency is actually a good thing to have and helps prevent unfair practices where (e.g.: men get more than women or nepotism gets you access to better salaries).

This isn't implicitly true. If it were, Iceland wouldn't have had to pass a law to make unfair salaries (based on gender[0]) illegal.

[0] - https://www.businessinsider.com/iceland-gender-pay-gap-law-2...

I mean. You should still make it illegal to discriminate, even if the tools to identify discrimination are available.

Otherwise what would you do with your discoveries? Vote with your wallet?

>This isn't implicitly true. If it were, Iceland wouldn't have had to pass a law to make unfair salaries (based on gender)

This of course presumes that any discrepancy in pay among genders is based solely on discrimination.

It's not everything but H1B anonymized salaries (not equity though) are public in the US: https://h1bdata.info/index.php?em=&job=&city=San+Francisco&y...

> Tax declarations are public information

What was historical reason for this? I see it as a privacy invasion. Is it a case for GDPR?

If it is public knowledge, then fraud becomes very hard. Also part of the egalitarian society that everyone should more or less earn the same, and nothing should be secret (at least to the gov).

That all your income and assets are automatically sent into the tax office by your employers, estate agents, banks make fraud even harder. But it does make your self-assessment super easy, just mostly log in and approve the pre-calculated assessment.

The digital transformation 15-20 years ago did have some issues initially when it became super easy to find out the income for all your neighbours, family and friends. It was open to all for a few years in Norway.

I have to admit I could not resist the temptation then as well. They have subsequently tightened access. You can request to see someone' details, but you have to be logged in and they will be notified who viewed it. I think journalists still do that to view celebrities details every year as there are always articles in April about their earnings.

Thankfully all the anonymised statistics are still public knowledge so you can easily compare how your earnings are with other people in your age group in your area etc.

You think you can still see a top 10 of whom paid the most tax in your area, had the largest income, or have the most assets. Handy for some more dubious people, probably.

My view is that public tax information is a good thing, on the whole. It creates greater transparency and harder to hide discrimination. It works in a society that is quite egalitarian, low crime and the public mostly trusts the government. It may not work as well in many other places.

The underlying reason is that we have a default that says that all documents produced by the different levels of government is public by default. This allows journalists and other investigators to expose bad stuff that is going on. Also note that just because something nominally is public, it doesn't automatically mean it is easily accessible. It can mean you have to go somewhere, submit forms and pay for the induced work load to produce the documents.

Things can still be made secret if there are actual reasons to do that, like privacy reasons or security reasons etc. Taxes have traditionally not been counted into the things that require privacy here in Sweden.

Socialism and strong unions basically. Your neighbor getting more than you is a big deal if you have strong unions.

Privacy in the Nordics is a bit weird. Everybody is issued a person number when they are born which typically starts with their date of birth. This person number is routinely asked for on all sorts of administrative forms and would also be on the tax declaration along with an address.

Before Finland, I also lived in Sweden briefly. When I turned 25 there, I received a letter from a local news paper (The Blekinge Länstidningen) asking me if I wanted to pay extra for printing a photo along with the happy birth day advertisement they were going to publish for free. They obviously got my name, address, and birthdate from the tax office and they did actually publish the ad (without the photo). Databases with addresses and person numbers were widely distributed. When renting a video tape, you'd simply give your person number and then they'd be able to look up your address. This was in 1999.

So, yes, I imagine there are some GDPR issues with that :-).

Here in Norway, it's for transparency.

In which country are they public.

"And there’s really no upside for individuals to upload and share their own private salary information, other than contributing to a grander vision of pay equality."

What a weird way for the author to end the article. Does there need to be a reason beyond pay equality? I benefit by having an understanding of the bands at a company. That only works if many people see value in this data and share theirs.

Yes, leverage is one of the most important parts of salary negotiation. If you know what a target is (i.e. what other people make), you are in a much stronger position.

For example, let's say you are just starting out, or are a couple of years in. You don't really understand how the business and hiring side works, you just want a programming job at a well known company. You might settle for 80%, 75%, 50% of what the market truly values you, because you don't really know what you can get. That's the important piece. It's hard to know your market value when you are a single data point while the hiring companies have entire teams analyzing the market for them.

This also compounds. All your raises over your entire tenure at a company are based on your existing salary. They typically won't think, "wow, we're only paying him $50k, we'd better bring it up to market rates." Managers don't get bonuses for thinking that way; they'll just give you a set percentage based on your current salary. You can start to see how important initial salary negotiation is; it affects your compensation for your entire tenure at that company. When hiring companies ask how much you make at your current company, they're trying to extend that discount to their company. All based on what you agreed to because you didn't have the pertinent data.

That visibility is so important, it's enshrined into US labor law (back when politicians gave a hoot about labor):

>Under the National Labor Relations Act, enacted in 1935, private-sector employees have the right to engage in "concerted activities for the purpose of collective bargaining or other mutual aid or protection."


I think you and I are in agreement. Sharing this data gives others power in their negotiations, and you benefit from others sharing. That seems like enough of a reason to me...

This is one of the reasons I hypothesise that you don’t get to hire good programmers in India. It’s not that there is a lack of talent. I’ve seen many enormously talented developers. However, the very best won’t work for Indian salary levels, since they know they can command better salaries abroad or become entrepreneurs. As a consequence there is an adverse selection effect in play in India. The very best won’t offer their services and the quality of the talent pool declines over time in a viscous cycle, as companies lower salaries even further to compensate for a poorer talent pool. Obviously this is not the only dynamic at play in the enormously complex Indian IT labor market, but it is a factor that I think that might be worth pondering.

Wait what? If you are a senior software engineer at a good company in Banglore you can easily make 40k - 60k USD an year. After taxes and expenses it becames equalent or better than salaries in most places except for San Francisco(USA), London and Singapore. USA is super hard for an Indian software engineer to immigrate. London is doable but for most of them it's not worth all the travel and staying away from family. People immigrate only if they makes way too much salary abroad compared to India which is hardly the case. Nobody is going to immigrate if they will save a few hundred dollars more a month. So good software engineers mostly immigrate for reasons other than salary (Like living in Europe etc).

Edit: Looks like even London don't offer good enough salary for a good software engineer when compared to cost of living.

London is fairly notorious for low SE salaries outside of high end consulting, I’m not sure why it would be listed with those other two cities.

The biggest problem with India is that talent migration doesn’t really work the other way (same with China), so you have to build companies solely with local talent. That works for lower end software services, but for higher end R&D generally requires some amount of draw from a world-wide labor pool.

I'm curious what's the difference between India and China in that it seems like there have been world-class R&D that has been done in China using local talent for a number of years now (from Alibaba, Baidu, Tencent, etc), but haven't heard of many examples in India.

None of those companies can really escape China, so is it fair to call them world class? Even tencent’s most international operations are acquired rather than grown in China.

I mentioned London because it had one of the highest salaries in Europe. I don't have personal experience myself apart from what I saw in Glassdoor.

Cost of living adjusted salaries are what matters, and London has SV level cost of living at way lower salaries.

In Europe, Switzerland is most similar to the Bay area: high salaries, high cost of living and low taxes.

Low taxes? California has high tax right?

Not compared to Europe I suppose?

Last time I checked take home pay was lower in NY than London for someone earning around 150k before tax. Not sure if that holds true in California...

According to this calculator: https://www.thesalarycalculator.co.uk/salary.php

120k £ (equivalent of $150k at current exchange rate) becomes net 71135 £ which is 61.8%

In the state of New York: https://smartasset.com/taxes/new-york-paycheck-calculator

$150k gets you a take home that is 68% of gross. NYC imposes an additional local tax between 3 and 3.8% so in the worst case you’re still taking home more than 64% of gross.

Maybe the last time you checked $150k was a lot less in £ and therefore taxed at lower rates in the UK?

Potentially- I think I added in property taxes (about 1-2k/year in the uk) and healthcare (free on the nhs or if you’re earning 150k you’re likely getting heavily subsidised private care from your employer).

Don't know many making $150k in London, though. Maybe the Facebooks, Google's

I think Texas market seems to offer the best of both world for software developers. Higher salary, slightly cheaper housing and cost of living. Also no state tax.

Yes, but: Texas weather. Unpleasant without constant A/C.

Indeed. Hell, I've known people from India to move back there in recent years because conditions have improved and whilst the salaries are lower the standard of living they can afford you can end up being a lot better.

The pollution in the big tech-centric cities though is still absolutely horrible, and this has to be a big factor for anyone with kids.

I wonder how many returnees were only able to improve their salaries by that foreign experience (warranted or not).

London salary to cost of living ratio is terrible. I made a better real income, and at a lower level, in Australia, which many people leave for better salaries in SV.

> Wait what? If you are a senior software engineer at a good company in Banglore you can easily make 40k - 60k USD an year. After taxes and expenses it becames equalent or better than salaries in most places except for San Francisco(USA), London and Singapore. USA is super hard for an Indian software engineer to immigrate. London is doable but for most of them it's not worth all the travel and staying away from family. People immigrate only if they makes way too much salary abroad compared to India which is hardly the case. Nobody is going to immigrate if they will save a few hundred dollars more a month. So good software engineers mostly immigrate for reasons other than salary (Like living in Europe etc).

That's not how salaries in a career path work -- you want to have the largest base to start. In a worst case scenario ( no job hopping for massive increases ) the salaries increase based on percentage of the salary, not a percentage of what is left.

Quite right. As a hiring manager I've experienced this first hand. You either have to settle for unprofessional engineers or those who accept the offer stick around, at best, for one year. I've seen engineers changing jobs within 2 weeks.

It's become really hard to hire good engineers in India at Indian salaries. It's actually good thing in the longer run if the overall salary levels rise. On the other hand the talent drain out of India continues to be quite high.

You are assuming that everyone hires only the best or need only best programmers. Starting with "good" and then in the middle you go with "best".

Stop rolling this lie about "only the best", there is a place for less skilled people. There is also no measure of "the best developers". Because of this lie, hiring in IT is broken, because people who belive that and are recruiting, shit on people who want to find a job.

It’s not that there is a lack of talent. I’ve seen many enormously talented developers. However, the very best won’t work for Indian salary levels

No, there is something more complicated going on. India has a space programme and nuclear weapons. They have the talented engineers to make that happen. I guarantee those engineers are not all “we are facing an issue” and “please do the needful” like everyone who has dealt with Indian IT has heard when they lack the basic competence to do the job. So where is the disconnect?

Working on a space programme and nuclear weapons is a "perk" on its own that can lower the salary. Maybe not for you, but for a lot of people.

It's like when people write entire programming languages for free. The job is so enjoyable that it lowers the salary to zero.

> It's like when people write entire programming languages for free. The job is so enjoyable that it lowers the salary to zero.

Dang, I thing you nailed it!

(I'd work on rocket engines for free, too!)

Are SpaceX salaries lower than average then?

Anecdotally, SpaceX pays less than other aerospace companies. Source: I used to work at Boeing, I have friends at SpaceX and various other aerospace companies.

How does SpaceX even succeed at what they do while paying below market rate?

It's supply and demand. Lots of people want to work on rockets, so SpaceX doesn't need to pay them as much. It actually is the market rate for rocket engineers.

Double down on this. I want to work in Aerospace, but not for the military industrial complex. Who do I work for?

You won't be doing any military work if you work at BCAC (Boeing Commercial Airplane Company).

Interesting. If true, this (and maybe to a lesser degree games developers) is a micro example of the macro job market. Akin to minor league baseball players, entry-level news reporters, bloggers, etc.

Not to mention the benefits that comes from a government job in India.

Well you get all kind of software Engineers in India. Software engineers that make less than 8kUSD per year to software engineers that make more than 70k an year. You get what you paid for. And most of the times foreign companies pay lower than what a McDonald's employee make an hour in USA and expect back services of a software engineer that gets paid 100K+ USD in USA.

You've just described brain-drain in its most common form. It's not a new phenomenon. It's at least as old as modern forms of transportation and probably way older than that.

A vicious self-reinforcing cycle that drains underdeveloped countries of the best and brightest - the very people most suited to lead and improve the countries they leave.

The US has also been historically the biggest benefactor.

Very true. Also, it is almost impossible to find senior engineer opportunities if you have lot of experience. Most companies I spoke to consider it almost a failure that with 15+ years of experience, I am not looking for upper management roles.

Here I am, still writing code every day after 45 years. One day, I'll figure out how to do it right, and so get promoted to incompetence.

I can second this. A friend of mine went through a real ordeal to change jobs as a senior IC.


This is a really silly comment.


It baffles me that someone could possibly think that in a country of over a billion people you can't find superstars. You'll find plenty, there will be a ~million one-in-a-thousand geniuses. Heck, if you take an average million people city and the smartest person in the entire city - India will have a ~thousand people that are smarter than that person, for any definition of smart.

In every thread: people finding reasons to simply bash Indians and looking down on them.

Just because of your interactions with few Indians and had a bad experience from those working in mega service companies like Infosys, Wipro doesn't mean everyone's a worthless engineer

Don’t forget one of the most talented mathematicians ever was Indian: Ramanujan. Of course , he was flown to London... as is precisely the actual topic “itt”.

Co-Founder of Levels.fyi here (mentioned in article). We aim to share salaries for all companies. You can see the salary bands for example of google as well here: https://www.levels.fyi/salary/Google/

Would love feedback and suggestions to help make this a better resource. Our next feature is regionalization. Look out for it in next few weeks!

Couple things I’ve thought of when using the site:

- Time at level

- External hire vs internal (promotion, role change, etc)

- As others have said, number of stocks.

As it is currently, it can be very hard to tell if a salary is representative of a new offer, someone at the end of a 4 year compensation “hump” due to vesting schedule, someone approaching their next promotion/high end of salary bands. To a certain degree the current version of the site suggests overall salary bands, but it’s hard to make it actionable for a specific case.

OTOH, I totally get that more data means more identifying, so tough balance. Love the site though!

It’s pretty hard (impossible?) for the market to reach an efficient outcome with asymmetric information. Sure, there will always be unobserved variables lid you indicate, but the opaque practices currently employed contribute to societal problems.

However, that same opacity gave me some peace of mind at a previous employer. I had very cutthroat coworkers. I was targeted, in part, for just the suspicion that I made more than other people. Didn’t matter if it was less than a quarter per hour. Didn’t matter when they all eventually were promoted above me and made more, either.

I don’t know how to square those two. On one side, market regulation ala talking to employees in search of abuses and mistreatment may have helped me. On the other is the persistent pressure to have the least government possible.

Thanks for the feedback! Regarding the first two, we collect Years of Experience along with Years at Company information. On the comp page if you click any of the rows, it will be shown. We have quite a bit of work to do on visualizations and will add this on the docket.

It would be helpful to have stock as number of shares. For example, Microsoft is up 40% on the year, Pinterest went public this year, etc.

Also, the graph from the Medium post showing comp vs years at company would be really helpful. In the OP, there's a 2nd year near $300k and 9.5 year employee at $150k.

Have companies every asked to be removed from levels.fyi? I see at least one company missing from there now.

Interesting, will add that to docket for when we release dynamic visualizations. One thing to note though: typically when folks switch companies they negotiate on the compensation they would be earning at current job (which includes the stock upside). This is actually why FANG comp has exploded. To attract talent whose stock has gone up, companies have to offer the same or more to poach that talent. Factor in a bull market for the last few years and it results in the massive offers we see today.

Re YoE Graph: Noted. We’re hoping to build dashboard functionality of sorts where you can choose what you want plotted and we'll dynamically generate graph.

No one has ever asked us to be removed. In fact, many companies have reached out to get their levels added. Which company is missing? We realize we need to add many more companies and are slowly adding more as we get data for it.

I don't understand why number of shares would be useful. levels.fyi asks for stock grant value, not stock vest value.

Number of shares is useful in several ways, especially for new employees looking to compare the value of a promotion / offer versus what the company is paying somebody 2-3 years their senior. You can backtrack and ask “what did the company think its stock was worth two years ago?” If a company drops the number of shares much faster than the stock appreciates, then that can signal they over-value their stock. Both Facebook and Uber saw large declines in the stock price during the 1-2 years surrounding their IPO.

Facebook was going for $40-45 during the months before their IPO (at $38), and 6 months later they fell to $20. Microsoft today might have a similar boom and bust. In these sorts of cases there will be lots of employees compensated unfairly one way or another.

When total comp offers are all similar, a major differentiator can be how much hubris a company shows valuing their own stock.

>especially for new employees looking to compare the value of a promotion / offer versus what the company is paying somebody 2-3 years their senior.

If you want to compare what you are given to what someone else is given, isn't grant price better than number of shares? Grant price is what the company gives people.

>If a company drops the number of shares much faster than the stock appreciates, then that can signal they over-value their stock.

Maybe. Or maybe the company is simply paying employees less. Or maybe they're migrating to provide more compensation via salary or bonus instead of stock.

Why would a company cut down on the number of shares granted if they overvalue their stock? Another thing is that a company could do a simultaneous stock buyback, which wouldn't cause the number of shares in existence to increase, and would only cost the company market price for the stock.

If you are given the average number of shares that an employee has vest per year, how can you determine if a company is dropping the number of shares? Really what you would want is data on every single grant event each employee has been given. If you know the date of each grant event, it's then equivalent whether you know the number of stocks of each grant event or the dollar value at grant time of each grant event.

Look at FB stock in the past 2 years and see how the same stock grant value has wildly different outcomes.

Past performance is no guarantee of future results. Just because some Facebook employees' stocks went up after grant, and some Facebook employees' stock went down after grant, does not indicate what your granted stock will do after you get hired by Facebook.

Also, if you do want to try to predict future stock moves via past stock moves, simply look at the stock charts. You can predict how much stock you will be granted via levels.fyi then adjust that based on how you predict the stock will move. I don't see how knowing how many stocks employees are granted will help you predict anything.

It would be great to add non US locations too.

I think in the UK it would be much easier. Outside of London, only businesses who have just done IPO will pay anything above the approx £40-80K for developers. Even Development Managers outside London are probably lucky to get more than £100K

On the other hand, we get healthcare paid for out of our taxes and accomodation is not too bad in most places so I guess it all works out in the end!

Also, businesses like Google and MS are poor calibrators anyway since they have massive profits compared to 99% of other businesses in the world and ironically, despite their crazy salaries, I don't think either are the best model of amazing software (has anyone used Outlook or GMail recently!?).

And filtering based on location, which affects the salary as well.

I assume you mean for visualizations. I do want to add though that you can already filter the comp data by location using the search bar or clicking the filter button to the right of it. Ex. https://www.levels.fyi/comp.html?track=Software%20Engineer&s...

Big fan of your site, the level comparison between companies is really useful. Keep up the good work!

Its interesting how completely different the numbers are on your site vs everything else I've ever seen.

Why do you think that is?

Example: https://www.indeed.com/cmp/Google/salaries vs https://www.levels.fyi/salary/Google/

Couple reasons. Some folks already mentioned below.

- Manual review - we review all data to filter extreme outliers and spam.

- By-Level info - we collect much more info per data point meaning we can hone in on title / role deeply.

- No Incentive - Not specific to Indeed but there is not incentive to submit your data. Everyone has access to data whether you submit or not. This along with the form being long leads to less spam.

- Staleness - we age out old data aggressively

- Total Comp - we're centered around total compensation. Most sites focus on salary which for tech jobs can sometimes be smallest portion of comp.

- Userbase - we're quite popular with tech demographic and have been able to field much wider range of data from folks at all levels due to it.

...there may be other reasons but these are the largest I can think of off hand

Not sure where indeed is getting their numbers, but I can assure you levels.fyi are well calibrated. Glassdoor also has weirdly low estimated salaries.

Has anyone done the sums of the numbers of people on l1->l9 vs the mean salaries estimated?

What does the estimated salary bill for each of these companies come out as and how does that stack up vs. their public declarations?

I haven't done that specific analysis but stock based compensation is given out in public quarterly financial statements. See here for example that Google spent ~9.35 Billion (!) in stock compensation last year: https://www.macrotrends.net/stocks/charts/GOOG/alphabet/stoc...

Glassdoor tends to underestimate because a lot of the datapoints are from several years ago.

Is it possible Indeed is only counting base salary, and not stock and bonus?

That's what I thought, then I looked at their estimates for Netflix Senior SW Engineer and found them to be similarly way , way too low. At Netflix, bulk of total comp is base salary -- bonuses and stock grants are just not done.

From talking with friends, the levels.fyi estimates are nearly perfect, indeed and glassdoor are way off.

Looks like your site is broken when viewed on an iOS mobile with Firefox Focus (anti-tracking) is enabled. Not sure why disabling the trackers should break your site...

We'll look into that right away! We keep track of which companies users click on so that we can focus data gathering on those. We use Google analytics so it's possible that's interfering.

Just as a historical data point: for a long time blogspot was completely broken if you blocked trackers (including Google analytics). The pages simply wouldn't load beyond the placeholder "gears in a box" emblem.

My hypothesis at the time was that Google had deliberately introduced changes to GA that made it far too convenient for sites to depend on on ga.js for providing [something in the path of] actual page rendering logic. And so for a number of years any URL pointing to blogspot was as good as dead to me.

The asinine behaviour has since been fixed. I can only assume the pageview numbers with intentionally parasitic ga.js were going down and couldn't be misattributed any longer.

Would be helpful to have salary by location break down charts.

Better than geographic would be state (counties for other countries) level break-down charts, as well. It would demonstrate the gross disparity between, say, Silicon Valley and DFW or Oslo and Trømso or Dublin and Cork.

I find these type of sharing salaries interesting since you never know if the people having a significant higher salary, want to share it. It might be a skewed way here high salary earners don’t want to put their salary in harms way and therefore don’t submit theirs to the system.

Therefore no real conclusive arguments can be drawn from this

Don't think it's fair to say that nothing can be drawn from it. Sure there's less statistical significance but I wouldn't discard it completely. There's a large altruism component involved. Nearly every other week folks email us (levels.fyi) willing to help us out with site development for free because we helped them negotiate a higher offer. If there's anything I learned from this project it's how willing people are to help others. Open-source is proof of this.

One thing that has always struck me about compensation is that it is incredibly infrequent where I can actually implement the type of fundamental changes a company most needs due to most opportunities for major throughput gains being hidden in the mires of interdepartmental communication breakdown.

I've actually gotten grudgingly bored of technical implementation/testing as at this point, I can look at a set of specifications, and generally have the code that underlies it it be almost immediately apparent. Migrations are a bit more interesting, however, at the end of the day, I get the most out of untangling or unobstructing human communication pipelines.

The work is irregular, generally novel, completely underappreciated, and nigh impossible to get anyone to agree to let you do regardless of the throughput increase you can enable. Especially if politics/an imagined threat to someone's budget becomes a sticking point somewhere in the chain.

So often times a job degenerates to the work an office will allow you to do.

It is one of the things I miss about startup culture sometimes. It's amazing how much a couple of people can get done with a process based on the right abstractions and a tailored vernacular/set of efficient protocols.

> One thing that has always struck me about compensation is that it is incredibly infrequent where I can actually implement the type of fundamental changes a company most needs due to most opportunities for major throughput gains being hidden in the mires of interdepartmental communication breakdown.

Then solve those problems. At higher levels, solving organizational problems is a job expectation for engineers. "Process is preventing me from having impact" is not an excuse.

Seems like the higher levels are a bit undercompensated compared to other FAANG companies but otherwise pretty average for a company like them

Yeah, it's an interesting trend, back in the day (day 10 years ago) it was conventional wisdom that moving to Amazon was a good move (financially) for more junior levels, but that more senior levels (64+) did better at MS.

From personal experience, the principal/staff band now pays significantly better at FAANG than it does at MS (considering total comp, because above a certain level salary isn't a useful comparison)

The stock compensation is just bad at every level.FG are far ahead.

No one's going to comment on the fact that when Microsoft employees want to collaborate, they turn to Google Sheets?

nice catch

Comparing US salaries to Indian salaries or anywhere else for that matter is like comparing apples to oranges. $50k equates to 35 lakh rupee per annum which is considered a great salary in Indian standards.

I know of many good developers in India who earn less than this figure but are generally very content in their life. Stop trying to look at everything through the lens of silicon valley.

I’m not sure “leak” is the right word for this, it looks like employees are voluntarily anonymously sharing their salaries.

It sounds like they agreed to share it with each other but not necessarily with the general public. So leak is appropriate for sharing it with the journalist.

Damn all these salaries in the 200-250k range. I worked as an engineer in the bay area for 12 years and never made over 150k or so. Guess I wasn't that good or didn't know how to play the game right.

I think the game is just leetcode and interviewing regularly at large companies.

Can someone copy paste. Medium pay wall blocking me

Just open the article in a private tab.

Can someone here explain why companies pay only 40% of US salaries in India? One person from my team is leaving for India soon. He is paid $200k/yr currently in Bay area. But he is being offered INR50 lakhs which is roughtly $80k USD. What stops MS from paying something like $120k USD (converted to INR) because fully loaded costs in US (with FICA taxes) is abot $250K?

Especially, if the quality is the same in both places?

Assuming $80k is market rate in the new location, if this person was paid $120k it creates another kind of risk: the person might be disgruntled and hate the job, but never leave, because they aren’t going to get $120k anywhere else.

As an employer the ideal is to pay at or slightly above the market. Enough to make money a non-issue, but not so much that people who secretly hate you still work for you.

Supply and demand.

MS could afford to pay more as you said... but why should they? Does your friend seriously consider leaving MS in India? If so, use that as leverage to negotiate. If not, then clearly MS should not offer more money

Not just India, tech salaries are pretty low even in the UK, much of Europe (except Switzerland) and even next door in Canada.

This just seems like a really complicated, roundabout way to do something a union would do automatically.

A union seems like a really complicated, roundabout way to do this.

What would be an easier way?


Every business is about "shareholders", the owners. That is why a business exists. Otherwise the names are charity, NGO, group of ferrets.

Please don't derail the conversation with unsubstantive comments.

Shouldn’t these kind of sites also include a productivity indicator next to each individual salary? Considering how much individual productivity varies for IT workers knowing just the salary is like knowing the price of a bag of, say, flour in a supermarket, without knowing the volume.

Also, to push the analogy forward, how about the flour quality? One employee could be more valuable for the business due to stuff like his willingness to help in tough, boring but critical moments. Or learning more outside of work to be able to apply that knowledge in new and innovative ways inside the company.

And so on... Are all those factors sufficiently captured by a single number like "level"?!

How would you collect productivity? I think that's quite subjective. As you mentioned, 'Level' is a signal of an individuals' performance to some degree. If you're performing well, you will move up quickly (broad statement but holds true probably in >80% of cases).

Perhaps performance reviews ratings? That should already get factored into the bonus at least.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact