I see parent's point as how you'd evaluate the performance of an international superstar in your team.

You'd do reviews where they'd have top performance, but not 3x or 4x more than the guy right below them. And if their presence didn't push your team to be #1 in that season, you could have an argument to replace them for someone cheaper and 90% as good.

Of course, their sheer presence could be a booster for marketing for instance, and there's a lot more going on, so usually you'd keep them around even if on paper it doesn't make sense, but it would look like a complex balancing act.

I somewhat get that, but if you are promoting people for amazing hits, then you are doing it wrong. You promote people not for being lucky and on a successful project, but for being a solid addition to the team that provides steady progress.

It's tough to distinguish between luck and skill. A lot of cognitive biases come into play in performance evaluation, and managers have to make decisions based on very little data.

> You promote people not for being lucky and on a successful project,

At Google you are explicitly rewarded at L5 and up for picking the right project to work on. It's wisdom of the crowds. You pick what project you're working on and you should be picking a good high-impact one that actually brings value to the company, for which you might be promoted. This ensures that incentives are aligned.

Google doesn't want experienced, highly capable engineers sitting around wasting their talents on dead-end projects, and they certainly won't be rewarded for it. You need to be proactive about ensuring that you're working on something valuable.

So Google Feed, Chat, Hangouts, social networks are out because they're highly likely to get killed.

It's so obvious for Google.

And that's probably part of what's happening at Apple as well ? We've seen important, but not quite groundbreaking products get left for dead, and eventually killed when at Apple's size it probably wasn't that problematic to keep them alive. I'm thinking about their network router lineup for instance. Or how nobody cares enough to make a Calculator app for the iPad.

> Or how nobody cares enough to make a Calculator app for the iPad.

Supposedly Steve Jobs hated every iteration of the calculator on the first iPad, so they just didn't ship it at all. Might just be tradition at Apple to not touch that topic.

I think its the opposite actually.

Google pays over $300k a year for many of their engineers, and Apple pays quite well too. Regardless of how easy that calculator app is, it'll take a team of people to build, and that team is expensive, like millions of dollars expensive, and it'll never be worth it.

>Google doesn't want experienced, highly capable engineers sitting around wasting their talents on dead-end projects, and they certainly won't be rewarded for it.

It's supposed to be management's responsibility to pick projects that will be successful. If a project turns out to be a dead-end, that's either management's fault, or just the way the market worked out and no one's fault (you can't always predict what's going to be popular or successful). And if a project is clearly a dead-end, then why is management wasting money on it instead of canceling it and moving the talent to better places? Basically, Google's system just rewards people for being lucky.

Engineers: "We want an engineering-led organization!"

Google: "OK, here's some real responsibility over company direction."

Engineers: "No, not like that."

I mean, I work at a normal-ass medium-sized company and at our top level of IC technical staff even we're expected to help management choose between different strategic options and point out places we can improve the product based on technical synergies or easy wins they may not be aware of. If we do a bad job at this, we will also get laid off on way or another, at latest when the company runs out of money.

Especially at Google's size, you can just throw up your hands and say "oh well, the market is hard to guess." You are the market!

Except Google is everything but an engineering-led organisation.

Google: "Here's some real responsibility over company direction"

Engineers: "So we get to work on things we want to move the company forward with, yes?"

Google: "Yes"

Engineers: "And we get rewarded for making good products that work well, yes?"

Google: "Only if the MBA clowns decide that's what they want"

Google's behavior is even more insulting than being purely management-led.

> Engineers: "And we get rewarded for making good products that work well, yes?"

This is not what engineers say (nor what the company needs unless you take vapid definitions of "good" and "well"). The MBAs aren't the ones asking for a dozen messaging apps!

Imagine paying someone 3-600k a year and saying Big Boy Boss Bob needs to decide what you do everyday.

Let the creative people you pay a lot make smart decisions and incentivize it.

> It's supposed to be management's responsibility to pick projects that will be successful

If you just do what management tells you at Google you will never be promoted past L4. You need to come up with new ideas and projects to sell to management. The key is to align your project ideas with what management says they want.

sounds really exhausting... if that's the constant expectation and you're actually good at it, why work for Google at all and not fundraise for yourself? cuz you're doing all the same legwork but your ideas go to the man.

If you can make 600k a year, a few years in a row (let's say three). How does 1.8M in hand stack up against starting your own business, with maybe a bigger potential number, but a lower statistical chance of making it there.

600k gross is 340k net in California if you are married.

Hopefully your spouse is earning a similar amount too!

Quite a lot of L9s did start companies, which were then acquired by Google. Others are working on things that already require a large amount of capital or influence, or just otherwise want to keep working on projects rather than administration.

It's not "really exhausting", it's just doing your job?? Personally I enjoy more coming up with my own stuff and fighting to launch it than just following orders (my favorite career accomplishments are making .app and .dev happen).

good for you, i can't imagine myself being under the gun to come up with a new funded project every year (or however often), would rather just work on some real product with revenue even if it's not very original

Based on this comment it seems like you have a highly distorted view of what it's actually like. You're arguing against a strawman.

It's not like that at all.

Also like, projects can last more than a year. There's something finishing up now that started work in like 2019 and planning in idk, 2016. And I know people who have gotten multiple promotions over growing their contributions to that particular endeavor.

I think it is exciting actually. Google a bit of a mess internally so there are lots of problems to solve. Everyone knows that these problems exist but people either aren't brave enough to defy management spend to the time to solve the problems or don't know how to promote their proposals to whom

If you are happy being handed pre-scoped work, you can just stay at L4, which many people do.

> It's supposed to be management's responsibility to pick projects that will be successful.

That's simply not how it works at Google. Everyone has skin in the game. Who are you to say how it's "supposed" to work there? I'm telling you how it actually is.

You get paid for being a solid addition on a team providing steady progress. If you continue to do the same thing and make the same valuable but incremental progress, why should your job role change?

This is one of the advantages of having reasonably wide pay bands at Google. Doing a great job within your level will earn you raises. You don't need promotion for that (to a limit). Why should promotion just be something earned after doing the same thing for a long time?

The parent also says "what did this person do to redefine the industry this cycle?". How often are you expecting a person to redefine the industry?

I assumed that was hyperbole.

Well it dilutes a conversation when someone implies a senior Google employee is surely to be expected to be at the tier of disrupting entire markets in order to get and keep their job.

If that is what really what defines an L9 job at Google I think we'd all have been praising the typical L9 Google employees on HN long ago. And we'd all be justified in being angry Google decided to fire such a person.

Or maybe Google should pay their salary indefinitely because they were capable of disrupting a market at some point in the past? Or maybe yes, such a person needs to work at Google type company so all of us can benefit from the optimal situation from which to maximize their talent, even though such a company would let them go. Otherwise we'd all lose out on such a person being capable of contributing to the world if Google lays them off.

What else could they do then without a company who doesn't appreciate them keeping them on?

I’m not sure what exactly is a L9 at Google, but this looks a lot like research positions in many huge companies.

For instance there was researchers on voice recognition at Yahoo! way before the voice assistants became a product, and if I could imagine one of them bring up a well working model 6 months before the competition, and have a voice assistant hit the market.

It would disrupt an entire market.

But they’re not alone of course, and I don’t know how many other researchers are on the payroll trying to breakthrough other fields. All of them have a “disrupt the entire market” expectation I think, and of course their employer is also expected to give them enough time and leeway to make progress. They probably bring small improvement on existing things, and the number of papers published every year could be another measure as a temporizing strategy. But their actual stated goal is probably very ambitious and they’d be failing at it for a very long time.

Vint Cerf

Not really sure what you're trying to say, but Vint Cerf is a VP (i.e. L10).

Research is a career path, not a level.

As levels go up, each new level is a new job.

In the SWE track:

L3 is a new graduate

L4 is mid level (an engineer who you can generally trust to complete the tasks they're assigned)

L5 is Senior (who owns/shapes the direction of the team)

L6 is Staff (who has to see upcoming problems and shape efforts across the organization)

L7 shapes a product

L8 shapes a division

L9 shapes an industry

The expectations are sky high, but so are the comp. If you want to get millions, you'd better do something worth millions.

Huh. So L7-L9 are L3s that are good at marketing and slap the Google badge on everything they do?

Well, I guess that explains both why they fire them and why golang.org became go.dev.

From your ranking, L9 seems to not depend on the company anymore, looks more research. You cannot expect to shape/redefine industry every year, that's absurd. By this model every math/physics/chemistry research teacher should be fired every year.

And to be honest I don't think google has these expectations. Most probably he was was fired for different reasons, not performance

Research is theoretical. These roles are very much practical. These are the kind of people who drive partnerships across companies, lead important industry standards, and come up with ideas for entire new markets and products.

This is not a job everyone wants or is capable of doing. There are a very limited number of such positions, and those who want them must justify the position's existence.

I can't speak for Google, and Amazon doesn't have an L9, so our Distinguished Engineers are L10.

But a decent bunch of them are the types of folks to have wikipedia pages:

* https://en.wikipedia.org/wiki/Jim_Roskind

* https://en.wikipedia.org/wiki/James_Gosling

* https://en.wikipedia.org/wiki/Tim_Bray (not anymore I guess)

The ones that don't probably would if it was public about how much of a significant contribution they've made to some foundational AWS services everyone knows. I think they don't mind.

Do they "redefine the industry" every cycle?

That's basically in the job description for the highest levels though. They're expected to literally lead the industry in their field. If they're merely now a solid and productive programmer/uber tech lead then they don't remotely merit such a high position. Since demotion doesn't exist for various reasons, the only solution is separation.

a lifetime ago back at Microsoft, Peter Spiro described these folks job expectation as "walk on water". if you wanted to exceed expectations, you had to "leave no wake"...

The very best engineer I've ever worked with at any company (Anna B.) just got promoted to Distinguished last year. She accomplished one or two industry-leading things before that :-)

Leading the industry clearly isn't good enough. They are expected to regularly redefine the industry.

Happiness is not enough! I demand constant euphoria!

Their managers earning as much or more don't lol

haha their managers' job is just redefining expectations of what is required.

A L9 manager has a very different job. That's a VP and they've probably got 500 people under them. Very high level ICs are also somewhat often directly managed by lower level managers (who are also paid much less). It isn't terribly uncommon to see an L8 managed by a L7.

Managing 500 people is a far lower responsibility than "redefine the entire industry". The entire industry is probably a hundred thousand workers and dozens of companies making hundreds of billions of dollars and you have to completely change the game for all of them. Why should someone who only manages a mere 500 people get more than one thousandth of the compensation? They probably only have a dozen direct reports anyway.

That's just a reflection of the fact that it's much easier to have a big impact if you direct an entire org to make it happen rather than trying to be a hero and doing it all individually.

We already know that part. They take credit for other people's work and they will fire you if you don't earn them a big payday.

> Since demotion doesn't exist for various reasons

maybe not in faangs, but I've seen CTOs and directors downshift to ICs in multiple startups and public tech companies. Honestly kicking them out instead sounds like a huge waste, if the person can make solid contributions in a different, maybe less intense, role, why not let them?

At Google it is actually common to have directors or VPs who don't have reports for a while, but we are talking about people who are ICs already. For each management position there is an equivalent IC position at Google with similar pay. I believe L9 would be a senior director.

From what I've seen these senior people with no reports at the moment (or those with org sizes too small) were hardest hit in the layoffs, simply because of the misfortune of bad timing.

It sounds like you're describing people getting a different job at a different company? That's not the same thing at all.

When I joined Google I was "demoting" myself from TL/M equivalent to mere L4 SWE. The much higher pay didn't make it feel like a demotion though.

nope, switching to a lower level at the same company

Well my origin comment was saying that it doesn't happen at Google. I can't speak for other employers.

Ah, clueless tech companies and the infinite growth myth, name one more popular duo.

Granted, it was hyperbolic, but at L9, often. This isn't just an extremely competent engineer, it's an industry leader.

I try to imagine the standups these people have: "Which industry have you redefined yesterday? Which industry are you redefining today? Do you have any blockers?"

It seems the right solution would be bonus-based compensation. If they redefine an industry, massive bonus, if they’re just a really good engineer, then get normal really good engineer salary.

Or a demotion. Is that ever an option in companies, or do they have to go straight to firing?

Google at least never demotes for performance reasons.

they fire. q.e.d.

Yeah totally agree. It’s basically Moneyball but for CS.

The Peter Principle is too simplistic. Some people do well and become incompetent over time. Or they stay competent but the company and industry change around them. It's possible that this person was laid off due in response to a brief period of subpar work unfortunately timed with a layoff mandate, and that would be terribly unfair, but I have no way of knowing that.

I took the anecdote to be about why these people would be let go even outside of layoffs. Layoffs, I'm willing to say, can and should be somewhat random. At best, they are controlled burns of a forest that you can't see. It will destabilize the ones that remain and you don't know how many of them you will now lose.

To the rest, certainly possible. My sports metaphor certainly implies a pipeline of workers that will replace each other in a somewhat predictable order. But that feels like a different thing than "they may be producing great code, but are not reinventing the world."

No one said these systems are smart. It's just the way it is and no one wants to change it for being the person that broke faang hiring. The alternative is a fundamentally different IC performance review process (and interview process) above a certain level, and HR and others are unwilling to take that step for obvious reasons.

L9 is conservatively making $2 million / year so the price of getting perf wrong is high.

Totally fair.

> The alternative is a fundamentally different IC performance review process (and interview process) above a certain level

What would that process look like?

Perf reviews these days are basically, "Did you do some important/useful things? Did you tell other people at the company about those things so they'd benefit, too?"

And interviewing (other than the inane hoop-jumping of getting through the recruiters, and the subset of interview loops that focus on LeetCode) is largely, "Can you write a bit of decent-looking code? Can you design a globally distributed self-healing system without being a total buffoon? Can you demonstrate some semblance of introspection?"

There's some bureaucratic fluff surrounding it all at the larger FAANGs, but it does seem largely reasonable overall — especially if you (or your manager) are willing to cut through the red tape.

I could see a fair number of improvements and tweaks to be made, but... fundamentally different? You want to throw everything out and do it all differently?

> The alternative is a fundamentally different IC performance review process (and interview process) above a certain level, and HR and others are unwilling to take that step for obvious reasons.

Having different review processes adapted to different job roles that exist at different levels is the easiest thing to defend. Heck, given disparate impact considerations and the fact most systems will have incidental unequal impacts, failing to do so could have you fall out of the business necessity criteria into unlawful discrimination at the levels the process isn't adapted to.

Are the “obvious reasons” that, despite the practical and legal benefits, someone would have to expend effort designing the appropriate system?

> This would be the equivalent for a pro baseball team laying off their entire roster because nobody on the team showed the growth that those in the feeder teams showed

It's been a while but as I recall this kinda thing is roughly the plot of "Moneyball". I don't think.they cut expensive players but they definitely didn't sign new ones who were expensive - instead they focused on players "who get on base".

>This would be the equivalent for a pro baseball team laying off their entire roster because nobody on the team showed the growth that those in the feeder teams showed.

This does happen in baseball, not literally the entire team, but Google also isn't laying off everyone. Teams regularly decide to let better performers leave just to replace them with worse players. The reason for this is that the worse players are always cheaper. Many teams are happy to pay a replacement player for half the production at a quarter of the cost. That L9 could have still been productive while not living up to the expectations for someone making that salary.

On sports teams (who plays baseball anyway?) you have a limit of people who can play at the same time. This does not really happen in a company.

If the expectation are easy to miss, maybe the salary of higher ups should be more strongly tied to meeting them. You can keep that half L9 at half of the price, while still making that worse guy that might improve over time.

Not just that but there are salary caps so sometimes trading away highly paid star is better for artificially constrained budget

Well, those are the expectations. Low ratings don’t mean being managed out in typical years, but that’s not where we find ourselves. I don’t think the system is perfect, but it is fair. An L5 that missed their expectations is laid off the same as an L9 - there’s no thought of “but they would’ve hit expectations as an L7” just as an L5 doesn’t get credit for being a high performing L3.

Yeah, from what I can surmise, I wouldn't be surprised if an L9 was earning salary equivalent to a top pro athlete (easily 7 figures). You better be writing some historic, market bending shit to get paid like that. Figure 10 years of a $2M salary, you've paid for everything money can buy, in cash, up to the private jet level.

> Figure 10 years of a $2M salary, you've paid for everything money can buy, in cash, up to the private jet level.

God, no. You're not even close to the limits of what money can buy. You may not even have enough money to be able to buy a private jet; it depends on the jet. You certainly haven't earned enough money that it would make any sense for you to buy a jet.

His point stands regardless

Yeah definitely no private jet. Maybe a little Cessna.

Assume that's $1m/year after taxes. With 3 kids in college and a nice house and a nice vacation a year, you're surely still comfortable but you're not buying a plane.

If it were my company doing layoffs, I’d hold on tight to the best and most senior engineers and lay off the entry level ones if I had to. Knowing what you have to do and why is much more important than the details of how you get there

It's a case by case basis and thinking you can apply a general rule across the organisation is completely misguided. I don't want to rehash the same arguments so just read the rest of the comments.

That's not really how it works. L6+ is considered a "terminal" level. As long as you aren't underperforming it's perfectly fine to stay at that level indefinitely. It's not an "up or out" position like L5 and down.

They keep raising it every time I hear about it…it used to be L4, now people say it’s L5/senior. And you’re saying it’s L6? I can’t keep up with this :/

I don't know where jedberg pulled that from, but I seriously doubt that the L6-equivalent is the terminal level at any FAANG. At Google it used to be L5, but as far as I know was dropped to L4 a few years ago

> It's not an "up or out" position like L5 and down.

As an outsider to this all FAANG thing that feels wrong to me, it has always felt so.

What happens, or at least what I think happens because of this strategy, is that lots of institutional knowledge gets lots pretty fast and in an irremediable way, because many of those L4s and L5s that ended up not making it to L6, and so got the boot, knew of some parts of the "system" that no-one else did.

Unless a FAANG company only relies on L6s and higher to get a hold over that institutional knowledge over the long term, which I pretty much doubt it can happen.

I don’t think L4 / L5 get booted that often, when was the last time FAANGS had mass layoffs ?

the more realistic scenario is attrition, going to another place for L+1 comp

As the median tenure at silly high wage places like Google is one year I am pretty sure most that quit are "managed out" with an implied threat of being fired.

At Google, this is wrong.

L4 is terminal. There is no requirement to ever be promoted to L5 (though there is some language about continued expansion of skills at L4). Most engineers will never make L6 in their entire career at Google and nobody is concerned about people chilling at lower levels.

> his would be the equivalent for a pro baseball team...

Where we're at right now, it's more like cutting Tom Brady. Sure, he's the GOAT, but they went 8-9 this year.

You can choose not to go for promo (at least at Google). A L7 is still making 600k+. If you've gone for L9 you are very aware of the ratchet and made a deliberate choice.