Hacker News Reader

Comment by munificent | Hacker News Reader

> We've proven that the big tech companies can go fully remote and not completely crash and burn, that's about it.

Agreed. We've proven that big tech companies can continue to be productive for up to two years going remote.

I suspect that that doesn't generalize very well to being completely remote over the long term. It works well when you already have a bunch of personal relationships between teammates that were established while the company was in-person. And it works well for experienced people that have ramped up. But I think it's probably quite a bit harder for people to be productive when they are new to the company, less experienced, and don't have that existing foundation to build on.

I think it's a solvable problem for companies that want to prioritize remote work, but I definitely don't think "managed to get through the past two years" means "it's a piece of cake".

View on HN

Luckily, there will be plenty of remote firms to scoop up all the engineers Google is about to hemorrhage.

What a miserable work culture it must be for them to even have the nerve to declare this.

Downvote away, I’ll be hiring some of them! ;-)

There is already a huge google engineer hemorrhage. A number of principal engineers and other folks I thought would be google lifers moved to snowflake. I don't expect this to affect Google's ability to continue to grow at 15-20%, though. There will be armies of younger people who still want to jumpstart their career there.

Google needs long tenure more than other companies due to its large reliance on in-house software. This is yet another move to shoot themselves in the foot. Luckily, network effects will keep them in business for another few decades.

Every large tech company relies on “in house software”.

Yes, but I can assure you there are vast differences.

Worked for many tier 1 investment banks in the techiest corners of each org and some cases felt like startups with all the shiny new toys everyone else was using vs the dark ages of customized software straight from the nineties or some other entirely-in-house-built abomination in other places.

Replacing senior engineers with juniors isn't sustainable though and their software quality is going to get a lot worse. Probably sooner than they think.

I mean... sustainably replacing senior people with junior people is the predominant arc of life throughout all human history. People retire or die.

You need to make sure you aren't replacing them so fast that the median experience level is dropping, but it's certainly the case that the average experience level of the incoming cohort is expected to be lower than the outgoing one.

It's not just about experience as an individual though, Google has hordes of arcane institutional knowledge about it's own platforms to maintain, like any large organization. Even if they were replacing everyone with senior devs and not juniors, you'd still be losing little bits of that each time someone left.

You see it at institutions, hundreds of individual "low bus factor" risks spread throughout the organization. The company would survive, but it would be a tough time, and a terrible place to work, during a large knowledge exodus.

Google has always lost top people to the next hip large-but-not-too-large tech company as they try something fresh / try to win the tech stock lottery. It was Facebook over 10 years ago, for example. You could get counter-offers to not leave for FB in that era. Is the current attrition any different?

Totally anecdotally, lagging comp and bureaucracy are much bigger sources of attrition than remote work. For the most part, anyone who wants to go remote can.

But Google let most people who asked to go remote, go remote. And said several times that if you haven’t asked to go remote yet, they’ll give you a 30 day notifications before you need to go back

I hate threads like these because people will circle around generic talking points while missing all the nuance and context behind this actual decision, like allowing almost any SWE to go remote that asks, giving a heads up, having a well-defined policy.

>like allowing almost any SWE to go remote that asks

Do you have the policy document or data to back that up? Asking because I am friends with a Google engineer who asked to go remote and was denied (even though nothing about their role requires them to be in-person).

A little birdie told me that most denials were for specific roles (eg some entire departments were not going to let employees go remote, mostly not engineering). In engineering the bird told me most denials were due to being very junior, working on something that needed physical access, or poor performance

What about being a new hire? Do they get generally approved or denied?

Anecdatum: a coworker who joined mid-pandemic was approved for remote work.

Thanks! But mid pandemic would be a year ago right? Not quite what I was thinking of as new hire, though cool to know regardless.

Yeah, but he applied for remote work when he was still pretty new. I don't remember exactly when, but only a few months in.

If you haven't been hired yet, I recommend checking the box for "remote-eligible jobs" and telling the recruiter that's what you want.

Full disclosure I work at Google.

https://careers.google.com/jobs/results/?distance=50&has_rem...

Seems odd to have the juniors and poor performers sit around in an office together while everyone else is at home

I wouldn't accuse Google of having an effective management culture.

They tend to promote from within and along the engineering track, which results in engineers reasoning from "What would I (a high-performing engineer) do?" and not necessarily what's best for the team.

There are other people who prefer to be an office that the juniors and poor performers can learn from...

The large majority of SWEs didn't apply for remote work.

Imo SWEs aren't on top of things like applying for remote work, and most who will, will do it on the last day they can

I find it very hard to believe that in the next month like half of the company is going to apply for remote positions. We didn't see this happen the last time we were within 30 days of a planned return (on Jan 3).

Orgs vary. There are absolutely people who had applications denied. All I can say is that in my org (hundreds of people), every single application was accepted.

OP's article says 15% of requests were denied, i.e. 85% of requests were approved.

I am a google manager of a hybrid remote team. It generally depends on your manager and your VP; both have to sign off. I’ve heard rumors that it’s harder for Junior members to get transferred, but I had two successful transfers, a junior engineer with a transfer request outstanding, and hired a remote worker today.

There is not a whole lot of nuance to the word "mandate". If Google was truly this amenable to requests to go remote, why "mandate" employees back to the office rather than make it optional? Sure, welcome people back who want to be in the office, but don't require people to be there.

Because the article reporting on it is shitty. Why not let the tens of thousands of Google employees who actually know what’s happening clear things up?

The people who applied and were accepted for fully remote aren’t the ones that need to go back. Only the people with an assigned desk in the offices that are re-opening are mandated to go back.

You’re getting tripped on the wording of an article that was hastily written after a reporter got wind of an email

>Only the people with an assigned desk in the offices that are re-opening are mandated to go back.

But you are still using the same word that I was getting tripped up on. Either the return to the office is optional or it is mandated. It can't be both.

The fact that the policy is company wide certainly lends more pressure towards returning to the office. That being said something can be mandated but poorly enforced. In the Google case, it appears that the enforcement is tuned down by an (apparently) usable exception mechanism. One can infer based on this, that in the language of the article, the new policy is somewhere between “mandated” and “optional”. With all due respect, I think you are being overly pedantic.

It can be, and is, both.

The decision to work as remote-first or office-first is, as I understand it, much more broadly available than it was. You are given the option to go remote-first and not return to the office at all. Returning to the office is optional.

If you choose to go office-first, then you will be expected to actually be at the office. In that sense, return to the office is mandatory.

The only thing that's really happening is them saying "you need to commit to one approach or the other, and act accordingly". This is understandable, you need to plan office capacity and other such practical considerations

It's team and role dependent, so it pretty much is "both" at the company level.

This seems win-win to me. Teams and orgs that consider remote-friendliness important can operate that way, while teams and orgs that don't think that can keep the team colocated. Ultimately, it means the vast majority of people's preferences are accommodated. Worst case, changing teams at Google is very smooth.

If it were truly "optional" for everyone even without manager approval, that's effectively "remote" because some people not being present means the team works in a largely remote fashion, and it also means space planning for expensive offices is very hard and result in office layouts that are not as productive.

For a big company with 100k+ employees, this seems like a good call IMO because it allows all working models to co-exist, and it can easily be tweaked after seeing how RTO goes.

Notably, the word "mandate" is not in the actual communication sent to Googlers. That word was chosen by the author of this headline.

Googlers with assigned desk will be expected to show up to the office. Applications for remote transfers have been available for like six months and in the cases where engineers are unable to work remotely due to their role or their management disapproving, transfers within Google to other teams that are remote-friendly are nearly trivial.

It is mostly orgs like Sales that have hard rules against remote work where the word "mandate" with its connotation of a hard and unpleasant rule makes sense.

>It is mostly orgs like Sales that have hard rules against remote work

That seems weird for outside sales (i.e. large accounts) given you're often near your customers and (in normal times) should be in customer meetings rather than sitting in an office.

There's still more nuance here - you can switch to remote, but will you get the same compensation?

No, you won't, just like you wouldn't if you switched offices. Google is and always has been very clear about this. They pay market rates based on location.

Because where your computer, desk, and chair matters of the quality of work you do?

No. These are shitty capitalist ploys trying to tell you the above is true when it's blatantly false. And well, they'd work you for no money, but they wouldn't have workers.

If anything, I'd claim to live in a high CoL area while being in the Midwest. They get their quality work, I get my commensurate compensation. Fuck them for thinking otherwise.

If companies had to settle on exactly 1 pay rate everywhere around the world, the last thing I’d be willing to bet on is that they choose the salary of their most expensive COL location.

You’re right. Let’s align everyone worldwide on the salary of Google Hyderabad

You sound very angry.

Your compensation is not absolute dollars but what you can get for those dollars.

Ideally the same dollar would get you the same everywhere in the world. That is not the case atm.

Instead, people at your new location will provide the same quality of work for fewer dollars, because they see what they can get for those dollars. That is your capitalist competition.

> Your compensation is not absolute dollars

Well, actually it is.

It has nothing to do with cost of living; it’s market rates which are set by supply and demand. London is super expensive but the pay is much lower than even the rural American pay scale since UK engineers don’t get paid well anywhere and will work for Google even if they’re getting half of what they’d get in the US. They could move to the US—Google would support the transfer—but they don’t, so there remains a supply of people who are willing to work for that rate in London.

As others mentioned, if Google ignored supply and demand and paid for quality of work only, why would they pay everyone the high rates that were originally set by supply and demand in Silicon Valley? They’d pay a normal living wage for someone in India.

They throw around "market rates" like there's a local market. Grouping employees into buckets by postal code makes sense for work-from-the-office arrangements.

On the other hand, in theory the bottom markets will see their rates rise quickly (because they merged into one wider remote market) if remote is as popular as it apparently is.

You do keep your compensation if you don't move your address, otherwise you get paid a market rate adjustment for that area, which from what I've seen is quite fair.

Yeah, market rates, but those studies end up in bad offers outside tech-hubs as top 80% of the market is still targeting way less skilled people than the one they "want" to retain. It's almost seems like the policy is designed in this way so it sounds fair and at the same time keeps new locations from popping up

>Since last June, Google has approved nearly 14,000 employees globally to transfer to a new location or go fully remote, Casey said. About 15% of applications have been denied, he added.

15% is still something I consider significant. Why were they denied? Are there people getting preference based on manager or their own personal reputation?

The most common reason is being a junior employee, who could still benefit from in-person mentorship.

At least until Q4 2021, Google also told large groups of employees not to bother applying for remote, as their request will automatically be denied due to a combination of role / team / organisation / tenure / office location / remote work location.

They may have reversed stance after I left, but I'm pretty sure the "15%" number cited is "15% of people that were considered eligible and had their manager support were declined."

Of course they'll lose some people but an office is enough of a competitive advantage to justify it.

Upvote you and more!

I'm not going back to the office ever there's no need to.. the company I work for has experienced explosive growth since we went remote ..brought on and maintained 100 new hires.

If they try to force me I'll find something else!!

Cities are dying some cause of remote work but aren't we tired of the govt controlling our lives (Biden's remark of all remote workers must go back..ha lmao). Umm the solution is to innovate cities and offer programs for remote workers to come live there for three to six to 12 months ..have a startup that owns buildings in all cities for remote workers to live in and congregate as well let's them jump from town to town every many months. City living Innovation not more govt control/lame ideas is what is needed to spur the economies of cities and their future prosperity.

This makes perfect sense, some people will be suited to some days on-site, and some off-site, but there will be others like myself where the commute is now a deal breaker.

If you get some talented engineers to join you because of the shortsightedness of Google then kudos to you, and more the loss for Google.

Every single thread I see on this has this exact same comment. It's like a bot wrote it. Is it that hard to imagine that an onboarding experience remotely can actually be done well, or that you could actually make friends with people existing solely on zoom? I think people are making that assumption, just because it feels right or maybe because their own org is just throwing juniors by the wayside and hoping covid will end rather than thinking about how to train people electronically from the ground up.

Remote work in a lot of roles, especially knowledge working, is just the way forward and will be how these things function in 50 years. Companies that are able to scout talent globally will simply out compete those that insist on a local labor pool. It makes no sense to perpetuate commuting, just in terms of the environmental damage it causes, when we've shown that this work can still be done without having to load a single ~200lb occupant and spend energy moving them + 3000lbs of metal around for two hours a day five days a week. If you are finding your juniors are falling behind, then step up instead of giving up and work hard to come up with a viable pipeline that isn't "well hopefully in two months we are back in the office." Plenty of companies and organizations and research groups have functioned entirely distributed for years now even before covid. It's not rocket science.

It's mostly a realisation most organisations are illogical, bloated, developer unfriendly and/or too undocumented.

We applaud ourselves for self-studying, being fairly educated on average with backgrounds in being taught to self-study effectively and finding information online. Yet somehow the idea of juniors doing those exact same things, with or without digital aid of seniors, is unthinkable to many.

Surely somewhere this story feels off.

Has anyone had a great onboarding experience remotely? How did it go?

I had a very well organized and documented remote onboarding experience. There was a list of meetings to schedule with the names of the relevant people and tasks to undertake, as well as a dedicated mentor that you met with as often as you'd like. At the end you met with at least one person on each team and got lots of time with your own team while they explained the systems to you.

Of course this is at a company that was remote before the pandemic.

I just wrote a sample of our onboarding process in another comment, and it sounds similar. And yes, we were remote pre-covid.

When a person signs in the first day, they already have a bunch of meetings scheduled (meet and greets, explainers, etc...), some work assigned (I like to have a bit of a goal when exploring a new code base, but this work isn't expected to get done right away), etc...

It can be done well, but takes some thinking through.

Is it that hard to imagine that an onboarding experience remotely can actually be done well, or that you could actually make friends with people existing solely on zoom?

We don't need to imagine it. Those of us who lead teams have tried, and we know it's harder to onboard people, and to socialize as a company, when you're working remotely.

I think the advantages of remote working greatly outweigh the disadvantages, and I will continue to work remotely if I can, but there's no need to pretend everything is better. It obviously isn't.

> I think the advantages of remote working greatly outweigh the disadvantages, and I will continue to work remotely if I can, but there's no need to pretend everything is better. It obviously isn't.

I think it's worth separating the two parties in this (the employer and employee). IMO too many comments focus just on the employee view and then wonder why the employer behaves differently.

Most employees seem to love it, although some certainly don't (I fit in the second category) - but most seem to find it preferable.

However my experience leading a remote team is that IMO it is worse for our particular company from a corporate perspective. On-boarding and knowledge sharing is worse, relationships degrade and it's hard to distinguish someone who is just struggling/slow and needs help with absenteeism.

I'd go so far to say that wfh is (generally) great in the short term, and awful in the long term.

Like, when an existing team goes WFH (e.g. when lockdown first kicked in), everyone already knows each other. You have your banter channels, you know who knows what about what, and for those conversations you'd prefer not to have on Slack you have non-company backchannels (e.g. WhatsApp).

But in the long term, after new people join the company and others leave, this degrades. It takes far longer for newbies to get to know people, social ties weaken, and trust falters. Doubly so for early-career joiners, who likely haven't worked an office job before.

But in the long term, after new people join the company and others leave, this degrades. It takes far longer for newbies to get to know people, social ties weaken, and trust falters. Doubly so for early-career joiners, who likely haven't worked an office job before.

That's all true if you let it happen. You don't have to though. If you recognize the problem you can put resources into addressing it.

> On-boarding and knowledge sharing is worse

I wouldn't say it's worse, but different. It requires thinking through the onboarding process in a deliberate manner. I think a lot of companies used to get away with onboarding by having the new person spend a bit of time with HR and then throwing them in with their team and hope it works out.

My onboarding process now is scheduled out and takes 1-2 weeks depending on the position. I make sure the new person has their computer a couple days before they start. I ask the new person during and after the process what's working/not working and refine the process from their feedback. I keep an open block of time every day for the person as long as they need it, etc...

I should add that we were already remote pre-covid, but still had to deal with all the same issues when we went remote. One of the biggest challenges we had and I see is that remote magnifies already existing issues around communication and management. I remember one of my managers asking 'how will I know if someone is working if they are not in the office?' and my response was 'how do you know if they are working now?'.

I'm also not going to argue in absolute terms that in-person or remote is better/worse than the other. They are just different.

The main problem with on-boarding people remotely isn't getting them equipment, it's the fact that lots of people don't develop good relationships with other people, and learning that would have historically happened by 'osmosis' sitting next to each other and casually chatting gets lost.

What if people don't want this tech pseudo-utopia? I understand where you're coming from and definitely agree that commuting is a problem but... we're already on the verge of a population crisis in developed nations, loneliness and alienation are at all time highs. I really really don't think that remote work is the utopia you imagine it to be. It's going to lead to more people who never leave their home or apartment, and maybe go days without interacting with people face to face.

We don't know that yet, do we?

For many, the exact forced socialization through work, commutes and office culture that's going on is what drains them to the point of isolating themselves afterwards. I expect most people no longer working under that system to find a way out of their social isolation, given the friction for those venues isn't too high.

Maybe if we stop pushing people into a catch-all environment and give them time, money and means to solve their problems, they will.. solve their problems. All of those three have been dwindling over the past few years.

>that you could actually make friends with people existing solely on zoom?

I don't get this idea that you need to be friends with your colleagues. Hpw does that improve productivity?

Professional yes, friendly even sure but friends? That's for leisure time and it shouldn't matter if their your colleagues or not.

There's a 95% chance that this is a disguised RIF.

People they want to keep will be able to continue WFH if they want. People who are on the margins or destined for the Performance Improvement Camps will have to come in and warm seats if they want to eke out a year or two more of existence.

This is quite sensible comment. I am sure not much liked here.

At my work at very anti-remote employer, there are so many people who have obtained permanent remote. So I wouldn't be surprised people who Google thinks valuable would get individual arrangements and other who just themselves think they are valuable are in for rude surprise.

What is an RIF?

Reduction in Force.

Reduction in Force. It's a term commonly used when companies are undergoing a large scale employee trimming. I don't think that google would do call it that. Probably just managers encouraged to increase the % of people who go on PIPs.

Probably just managers encouraged to increase the % of people who go on PIPs.

That's how tech companies usually do layoffs. Rather than admit they overexpanded and have to cut people (or that they are cannibalizing their own people to boost executive compensation) they blame it on departing workers by increasing the PIP-rape cutoff.

A different way to do it is to bring on contingent labor instead of hiring people as permanent employees. I don't know if that is better but at least people know what they're getting.

What are PIPs?

Performance Improvement Plans.

Basically a company's last effort to cover their butts before firing you. If you contest the termination, they can claim they gave you fair warning and a chance to improve.

If you're ever on one, the sensible course of action is to look for another job. You're next on the chopping block.

Reduction in Force - way to lower overall employee headcount.

Reduction In Force

Parent asked what it is, not what the acronym stands for. It's a modern euphemism for layoffs, redundancies or mass firings, sackings, depending on your flavo[u]r of English.

You can just fire people. If I had a report that was not meeting expectations I sure as hell wouldn't make "deny their remote work application, hope they are dissatisfied with the office, and wait for them to quit" be my plan.

In big tech companies, who gets fired has very low correlation to performance. Instead, it usually comes down to:

(1) people who make their bosses look bad are first to go.

(2) next are people who are perceived to cost their bosses time, regardless of whether it's their fault.

(3) after that, it's usually the infanticide cases: i.e., the people who did nothing wrong but haven't been there long enough to establish themselves.

The infanticide is especially ugly, because (a) it means the company is firing people basically at random, and (b) it rapes the shit out of the resumes of the people affected, because they now have <1 year jobs to explain. The reason it happens is that, empirically, most of the people cut in mass stack-rank purges are new members of underperforming teams... who, by inspection, have had the least to do with the team's underperformance.

Underperformance does get people fired, but rarely. It's at least as likely to get someone promoted, because underperformers usually have a career's worth of experience of being shitty, and therefore have developed such political skills they can easily fail up every time.

Exactly my experience as well

> You can just fire people.

Not at Google, you can't. Firing someone who's performing abysmally still takes up to 6 months, between all the process and PIPs and paperwork and shit.

Google's preferred method to "fire" a lousy programmer is to make him a PM.

In my experience, SWE -> PM ladder transfers are very unusual.

I'm sure they are. Most SWEs hired at Google are pretty good at programming and don't need to be tucked away in PM.

Xoogler here, from what I've seen, SWEs converting to PMs did it because it was what they were interested in it, unrelated to their skill, and it's not trivial at all. IIRC (never done it myself but seen a couple of people who did) there is a trial period and if you don't perform well enough as a PM you either go back to being a SWE or need to resign.

PM is a negative contribution role, so what does "performance" even mean? Getting programmers to accept Jirafication?

I guess you're just trolling, so have a nice life.

Sure, but that is way faster than "wait for them to quit."

I'm not sure there is good evidence that tech was that productive the past couple of years by any work they were doing. Rather, they were the only game left in town.

> Agreed. We've proven that big tech companies can continue to be productive for up to two years going remote.

This sentence amuses me a bit. Does that also mean that American Democracy has only been proven to work for up to 245 years? :-)

I mean.. yes? Except not even that, really, because the actual modern structure of American democracy is really closer to 100-150 years old (ie. somewhere between the country being broken in half and the 17th amendment introducing elected senators). But in the grand scheme of "humans having governments" it's pretty short either way, honestly pretty proportional to 2 years vs. "humans having megacorporations".

Well yes. But in that sense, the Democracy has gone through a full lifecycle of elections, candidates starting, candidates retiring etc etc such that all processes for doing it are tested. So you can have confidence that it can handle a lot of things. But not necessarily ALL possible things so it may well be that something comes along and kills it. Maybe in 10 years, maybe in 100000 years. But for now, it seems pretty stable.

OP's point is that 2 years of remote work isn't a test that a company can successfully go through a lifecycle of a team of employees. You know, onboard a grad, get them up to speed, get useful productivity, maintain whatever system/documentation, have them leave AND repeat this until the original members of the team are all gone AND then check that things are still ticking along well.

It may be that said companies are "going well" because there is 1 person left holding the fort while a bunch of new starters fumble around with their remote work, not getting up to speed properly.

Unlikely maybe? But the point is that 2 years isn't a proper test of such working conditions.

EDIT: remove political rub.

I think so. Systems like remote work, American Democracy are not simple enough to have a scientific proof to provide guarantees. So many factors that might not otherwise have affected the outcomes could affect an outcome in future. So the only fair statements are that remote work has been proven to work for 2 years and american democracy for 200ish years :-) This is quite fascinating

Depending on your perspective, it may have been working for considerably less time than that

The typical tenure for many developers is 2-3 years. 2 years is a long time for a software company.

That really depends on the software company. There are plenty of places where two years as about two releases of the product.

Indeed. Not to mention if you're at one of the bigger companies then you're more likely to change teams than leave the company entirely.

> I suspect that that doesn't generalize very well to being completely remote over the long term.

And when I say I am now working 11 years remote and was always more productive compared to be before (9 years in offices) you'll immediately say "anecdotal evidence", right?

Maybe look into your bias. It does look like you're trying to find scientific reasoning to support your subjective preference. That's not OK. Let's at least have an objective discussion.

Demeaning the positive results of remote work by pettily narrowing down their results to super specific borders, while at the same time implying that office work is the go-to thing to do is not being objective.

I joined a new org earlier this year, fully remote, and my last gig was only a week in person before the pandemic started. I have also onboarded half a dozen juniors in the last few months, again all remote. Your comment is completely out of touch.

My company is fully remote, and whilst I'd agree that it's certainly doable to onboard people remotely (we've absolutely done this successfully), I would also say it really depends on the types of people and I suspect possibly the roles as well.

>works well when you already have a bunch of personal relationships between teammates that were established while the company was in-person. And it works well for experienced people that have ramped up.

I think a key question here is how long and how often do you need to see someone to establish and maintain those relationships.

There were exporters/importers and othe long distance businesses relationships for a long time, so I don't think its every day for years.

  > It works well when you already have a bunch of personal relationships between teammates that were established while the company was in-person.

Clearly you have no idea how much churn is common in SV. At google after a 2-year tenure your "percent" (what percent of SWE is newer than you) reaches 50%. What personal relationships do all those who were hired remote and never saw the inside of the office have?

That is just insane churn. There has to be so much wasted work when so much project knowledge is dropped.

Two years is nothing. You just got warm in your cloths and started to perform.

Edit: There most be impossible to have a stable engineering "culture" in such companies.

This is made somewhat misleading by insane growth. Even if there is no churn your percent is still surprisingly high because of new hires that are coming from new headcount.

Reply on news.ycombinator.com