The levels.fyi data is a good ballpark for comp with value at grant (and it matches my experience, which I’m fairly sure was at least at the median for L6).
700k means either exceptional extra grants or a lot of stock appreciation for L6 (so value at vest being much higher than value at grant), and even then I don’t think it would be common, it’s more like comp at L7+.
Heavy, because the real money enters when you compound the returns and when you are able to stand the losses.
I even made less than her working when I FIREd, but not anymore. Poor girl for paying someone’s retirement with her Thalwil mansion.
Folks, I hate to be the one[‡] telling you this, but this is your own fault: you index
loversbuyers keep buying the VT and other market weight indices, continually and relentlessly propping up these companies which keep diluting their shareholders at break-neck pace (e.g. $22.8B in stock based comp in 2024 just for GOOG).
Where do you think this money is going?
Well, Alphabet had 62B$ net buybacks after stock compensation, so it isn’t that diluting overall?
Nice, but so what? Working with hoolie always gave you a cheese and ham sandwich for free…
The trick is not getting addicted to the money. Like with a potent older boyfriend who pays part of your thousands of CHF Thalwil mansion,..but covers your part too. Or learn how to do it yourself.
Disclaimer: I can only speak for the time at Google before the turn of the decade.
The manager determines the new total comp at grant time,[2008] the employee obviously sees their total benefit at vesting time (or typically a little later, as grants usually vest during blackout periods, unless the employee signs up for automated selling at the time the grant vests).
There is none in reality.
In fact, there was a period of probably 5+ years when it was “employee insider knowledge” that it was better to stay at Lx with “strongly exceeding expectations” – a rating indicating you would “meet expectations” at L(x+1) – as with enough insider knowledge you would just be “gaming the multiplier numbers”.
I (ab)used this myself for years.
The numbers to game were numbers set by the comp committee which would, well, calculate things to come up with (different) multipiers for performance ratings at each respective level.
Assume the base comp for L6 is 200k to 250k.
Assume the L6 is “strongly exceeding expectations” which – according to the comp committee – warrants a 1.2x multiplier.
Add in some additional formulas – remember, everything is a precise number to the 3rd decimal multiplied by a manager committee horse trading negotiated blurry bucket summary ratings that need to fit a curve – and you end up with calculations that end up with:
If you were a tenured L6 at 250k (base salary) compared to a fresh L7 at 250k you would do better as L6 not going for promotion.
Of course, Google corp noticed this and corrected things, but it only created different incentives, and as we all religiously repeat after Charlie (Munger): show me the incentives and I’ll show you the outcome.[1]
Looks like you weren’t on my team. 
To be fair, there were (and probably still are) different comp levels for different job ladders.
The way I read the tone of the article, she’s not thinking about compounding, but about enjoying some time off, with a nice backstop by a 17 year senior partner. Lucky him. Or lucky her?
There is not, by the way, an 18-month sabbatical with Google. There used to be a 6 month sabbatical that you could take after tenure of 10 years, IIRC, later reduced to a 3 months sabbatical, and then abolished completely: Google noticed that people wouldn’t just quit but “take a sabbatical” so their stock options would continue to vest even if their salary and part of their bonus was nilled. Given the equity was such an outsized portion of the total comp, “going on a sabbatical” before quitting was just too attractive …
Some (geo) jurisdictions addressed this loophole (for employees), but not everywhere.
Wow. So you approve of them buying back stock (and paying for employee grants) with the company’s shareholders’ profits?
Me, if I were a shareholder, I’d prefer to get the $62B distributed to shareholders instead.
I could then still make the decision to further invest into GOOG, but it’d be my decision instead of GOOG’s which apparently does not know better anymore than to allocate capital into its own stock.
2008 There was a repricing of then stock options (instead of the stock grants later on) because of the GFC. In the middle of the compensation cycle, Google decided to grant more options because the Google stock price deteriorated during the GFC in 2008 or so, to keep employees interested.
In my own ignorant recollection of things, nobody sane in their mind would have quit the company because the stock option grants halved or even became worthless, but who am I to complain about Google extracting a little more money from the market (from the perspective of an employee).
1 Admittedly, there were even more stupid performance indicators, like measuring one’s productivity by numbers of lines of code submitted to the repository.
For the non-coders: this is kind of akin to measuring a baker’s performance based on the number of kilos of flour going into the dough without taking into account how the bread tastes or how much of it is sold.
Wow. So you approve of them buying back stock (and paying for employee grants) with the company’s shareholders’ profits?
Huh, aren’t the employee grants an expense still (for GAAP accounting at least)?
And yes I think I approve of getting the profits out in a way that isn’t taxed in Switzerland. (kinda depends on how strongly I believe in EMH and Alphabet not buying super overpriced but …)
I’m talking about SWE ladder, where the main difference between org is mainly equity. It’s easy to verify given internally salary data isn’t very hard to find and I’m pretty confident about the fact that levels.fyi data is close to reality 
Did you say GAAP?
Glad you didn’t say EBITDA or I would have Charlie Mungered you …[$]
I’m fully confident about the numbers for the period before 2020 given first hand experience. For SWEs, PGMs, TPMs, Engineering Managers and Security Engineers.
Of course, your experience might have been or is different.
$ “I think that every time you see the word EBITDA, you should substitute the word ‘bullshit’ earnings.”
– Charlie Munger
If you want the story from the blog 
https://www.mustachianpost.com/blog/florence-quit-her-job-at-google-and-opts-for-coast-fire/
And another one:
https://www.mustachianpost.com/blog/florence-coast-fire-one-year-later/
They pay 6600 CHF rent for a lake view flat in Thalwil. I was always wondering who is willing to pay these prices. OK, these kind of people.
What’s weird is that she’s boasting about saving 80-85% of her salary, yet she retired with only $1.5 million, which is less than what I have. And I am still afraid to retire. At 3% SWR per year I could spend $200 per day, but what if I fall ill, what if there is inflation, what if I feel the need to inflate my lifestyle to stay happy?
Just date and friend a 17 year older gf that will take care of you because … love?
How could she resist your profile picture?
Edit: my profile pic isn’t exactly working out, either … any hot tips?
They didn’t retire from what I read, the call it mini-retirement (some people call it funemployment), she mentions looking for work again after her break.
It’s a good alternative to RE, take a 1-2y break once a decade so you can benefit from “retirement” while young and healthy rather than waiting until your 50s.
And a good excuse for jail time…
Effective hot tip: better start working out, then. 
Get paired quick hot tip: I have a non-refundable 10 part course in a secret Discord that, for the low, low price of CHF 10M per lesson can turn you into a magnet for older rich people with higher salaries credentials. There’s a serie for each gender, I’d follow them all to increase my chances if I were you. Drop me a PM if interested. 
Sold! Sign-up link?
They said 6.6k USD, which more like 5.3k CHF today. While that’s still quite a lot in my books, it’s much less insane than 6.6k CHF (1.3k less insane to be precise)
I’m curious if you could give rough numbers at where L6 comp topped out on your team. 800k? 1M? 1.5M? My prior understanding was that SWE comp is fairly uniform across the company, except sometimes for very niche areas where there is substantial risk of poaching that threatens business. Maybe you were in such an area?
in a similar vein, I’m aware that compensation amongst top AI talent can be a lot higher than “normal” SWE comp due to the current AI talent wars, but i understood that this is specific to the deepmind org which apparently has its own compensation process
It doesn’t sound as appealing when you frame it like that.
Now you made me feel sorry for the girl.
I know, I know, we’re being mean. I don’t care.
Thanks for the correction. Well, a monthly rent of 5’250 CHF at 3% annual yield would put the value of their flat at 2’100’000 CHF, which isn’t a bad price if you want lake view. I guess we’re faster to dismiss 5000 rent as crazy than 2 mil selling price.
Can’t speak to AI as I am mere human intelligence. My field was information security which was somewhat hot for a bunch of years after Google got breached in 2009 by China and Sergey then decided to do the right thing versus obeying shareholders.
Those times are are of course long gone, btw, if anyone was still wondering.
L6 comp in what I was able to overview would probably max out at now CHF 800k (then maybe USD 900k to 1M) before you’d be essentially “forced” to go for promotion.
Probably corner cases exist that I am not aware of that pushed this even higher, coasting on their “strongly exceeds” or “superb” rating while just not going for promo. I know of one guy who managed to do this, but he was not on my team.
SWE comp was indeed fairly uniform for “standard” SWE jobs. Some areas do require a little more “in demand” knowledge than your run-off-the-mill coding chops, but those higher standards would typically only trigger when run-off-the-mill SWEs would try to transfer into specialist SWE roles, essentially barring them from successfully transferring.
The (then) specialist SWE roles would officially not have higher level salaries, but inofficially were rated “beneficially” by managers (see my previous remark about the horse trading around grading them on their corresponding ladder) so the “more special” SWEs would just benefit from more quickly progressing on their ladder.
As I write this, I slowly shake my head about how accepting I was of this … stock market, really … claiming then how fair and balanced things were all the while trading my team (my stock) vs the market (the promotion committees I participated in) vs the outcomes of Plato’s allegory of the cave when I would be communicated outcomes I could not support and had no real influence on but that I would have to communicate to the report with the message the company wanted me to relay.
Anyway, looks like the 45 minutes of today’s therapy session for me are up already?
Different question: what would be the reason for her (you if you’re reading and want to out yourself here) to want to go public with this?