A few points, based on my experience at Facebook and Google. Facebook and Google use a surprisingly similar job title system (maybe Bret Taylor copied the Google system while he was at Facebook?) When you are just hired, you are usually a level 3 but maybe a level 4. After 3-5 years of experience, it is expected that everyone be promoted twice to level 5. Level 5 = "senior software engineer" so that is what Glassdoor will tell you about.

So, there are also plenty of levels above level 5, and it is not uncommon to get there with 5-6 years of experience. If you are world-class famous in your area, you might be a level 9 engineer, even though your external title will still say "Software Engineer". If you are the top engineer on your team you might be a level 6 or 7. Each level is approximately a 30-40% raise on the previous level. So one thing that's happening is that Glassdoor just isn't reflecting the upper end of what you can make.

Another thing is the value of stock. Stock can vary a lot from person to person and company to company. Some companies will give you essentially zero stock. Some will give you stock that's worth a comparable amount to your base salary. More common is somewhere in the middle. Obviously the value varies depending on how the stock does while you're working there, but there are also various ways to get extra stock bonuses for different reasons. In particular you can get stock bonuses for high performance, and if you are being recruited for your particular expertise you might also get an extra-large stock offer.

> world-class famous in your area

s/area/specialty/

I've noticed that as well. The problem is Glassdoor very rarely actually includes stocks (though they say they do ¯\_(ツ)_/¯). All my friends making $400k+ make more than 50% of that in stocks.

As individuals, we look at the world through a microscope, so all personal experience is anecdotal. But these $400k+ earning SEs are a double-digit percent of the engineers I know with 6+ years of experience (N > 10).

Can you elaborate more on "make more than 50% of that in stocks"?

It looks like GOOG is up about 120 points compared to this time last year. Is that $200k / year just from projected stock growth / dividends / etc? Do employees in this situation get issued $200k worth of stock every year? Or is it done via options? Is it a case of stocks vesting over time?

I find all this fascinating because it's so different from my experience. I work for a software company in the Philadelphia suburbs, and I have about 15 years of experience (all over the place - C#, Java, some C++, JS, etc.). When I compare myself to my local peers, I feel like I'm a pretty decent developer.

$200k+ / year total compensation seems absolutely massive to me, even factoring cost-of-living adjustments. Like in the realm of "too good to be true" or "results may vary". But now I'm wondering...

In most places you'd be happy for a Kwarter of that.

> these $400k+ earning SEs are a double-digit percent of the engineers I know with 6+ years of experience (N > 10)

Well they would have to be, wouldn't they?

All the anecdotes we hear about high compensation come with a heavy dose of survivor bias---we only hear about the high salaries. There are lots of good devs making more modest figures at less glamorous startups or small shops. Glassdoor reflects that.

The problem is that "software engineer" encompasses both the architects who design skyscrapers and the construction workers who install drywall in them

I have a lot of engineer friends in silicon valley making $220-280k, anecdotally

This would all be much more useful if the salaries were stated in some form that's normalized for COL. Maybe it could be factors, like earning "2" means you earn 2 times your COL. Or at least just earnings minus COL.

People really overestimate how much COL affects how much you earn in San Francisco. $70k in Atlanta is not close to $235k in San Francisco (both typical new grad starting salaries). Assuming you spend $40k on rent a year without having a roommate in San Francisco you still have $101k after taxes.

> $70k in Atlanta is not close to $235k in San Francisco (both typical new grad starting salaries)

Say what? Not even Google starts people at a 235K salary. Total comp package sure (so starting bonus, bonus, RSU grants). In the last year or so, the new grads I'm friends with are averaging around ~150K a year starting base salary from the big companies in the bay.

> Say what? Not even Google starts people at a 235K salary

Yes, that was my mistake. $235k is a total compensation.

I agree completely. Cost of living is dominated by your housing expense, so to scale you're basically just looking at the net delta in rent or mortgage divided by the tax rate, so like current_salary + rent_delta/0.6 seems a reasonable way to find a break even point.

Minus makes more sense to me, because you can save up and then move.

I know someone who turned down a job in Cupertino because they were only going to make about 20% more than Seattle, which would have effectively been a huge pay cut for a little more upward mobility and some prestige. The glass door numbers you cite are within margin of error of those numbers.

Is this just people throwing around VC money pretending it will attract the best of the best?

Glass door data is stale and I believe that’s only largely true at large companies unless you do real gymnastics with your stock options. It is true though.

I personally know people with < 3 years experience making $300k+. I know others with 0 years experience getting $210k+.

This is in Boston, Seattle and Chicago, not SF.

What.

A junior level making >200k?

Please elaborate.

What’s to elaborate? If you do well in your interviews at top tech companies or get a $75k signing bonus after you intern it’s very common.

It makes no sense to include a one-time bonus as part of one's annual income, unless one is trying to brag rather than inform.

It absolutely is considering that can be invested with a stable ROI. It still means more in a year than your peers without signing bonuses.

I'm not suggesting that anyone should turn down the money, but it's unhelpful to include one-time payments like a signing bonus when comparing salaries across the industry.

> Is the Glassdoor data bad

it is, use https://www.levels.fyi/comp.html instead

This is most accurate for mid-band TC specifically.

I think Glassdoor data is bad. Check Blind to see what people are actually getting paid.

It’s common for engineers here to get a significant portion of their pay in stock. Stock, of course has a tendency to change value, and in a historic bull run if you count stock growth you can get to some very high numbers for total comp.

Almost half the compensation comes from stock grants, which Glassdoor doesn’t measure. Look up what people have entered in the levels.fyi spreadsheet if you want to see actual comp.

Those values are almost always inflated upwards and should be taken with a few grains of salt.

I recently did a lot of interviewing and found the levels.fyi numbers very accurate to the offers / potential offers I was seeing.

They also depend heavily on the performance of the company in question. Amazon's stock went up 8x in the 6 years I worked there, so the original numbers vastly underestimate the reality.