This is a meme that people love and there's some truth to it, but also a number of ways in which it's inaccurate.
Sure you can get promoted for launching a new thing, but you can also get promoted for fixing a first pass implementation, for scaling something from a small user base to a large user base, and so on, as long as you can show the "impact" of the change – whether that's revenue, savings, productivity, improved UX, etc.
It's possible that up at the top there's more of a focus on shipping new things, but I don't think that's something that 99% of Googlers need to obsess over, and it doesn't reflect the bulk of the work actually being done.
The problem is that "impact" is easy to establish for shiny new things, whereas for preventing problems – you know, maintaining things – the impact is how much better you have made the universe compared to what the universe would have been without your work. Of course, you have done your work, so we never get to observe that other, "counterfactual" universe that establishes your actual impact.
To estimate the impact, we need to run controlled experiments. And we have lots of A/B testing infrastructure to help us run those experiments when it comes to launching shiny new things. But when it comes to maintaining things, not so much.
So if there are two possible projects of equal difficulty and impact that you could take on, and one is to launch a shiny new thing and the other is to maintain an existing thing, you're always better off launching the shiny new thing because the cost of quantifying your impact will be low, whereas had you chosen the maintenance project, quantifying your impact would be prohibitively expensive.
And that's why building shiny new things dominates maintaining existing things: you get full credit for your work, not a small fraction.
I think you've missed my point. As you break down "the grind", it breaks down into projects like "fix this bug" or "speed up that API", and those are impactful and can lead to promotion. Sure, no director is going to be promoted to VP for fixing a bug, but it's useful stuff for a new grad to have completed and would feed into their promotion.
If there is "grind" that is truly not impactful, then why is it being done? I'm not talking about for promotion, I mean why would anyone make changes that have no impact?
Because, especially now, we can see lack of it drastically impacting new Google products like Stadia. No one can trust Google to maintain their products at all.
Stadia failed as a business, not as a product or technical solution.
I work on Google Play, we've been going for a decade, no sign of stopping now. Much of what I and the many people around me do is maintenance – making things go faster, making things error less, etc. Some of that is done with new features, some is transparent to users (or app developers).
I don't see this culture that HN seems so convinced about. I can see elements of it at a very high level, as I said I'm not really commenting on the top leadership, but for most people at Google, new products do not appear to be necessary for progression.
>> Stadia failed as a business, not as a product or technical solution.
Stadia failed as a business because it failed to get adoption by developers and gamers / users.
It failed to get adoption by developers and gamers because Google has a reputation for killing things.
Google has a reputation for killing things because it's leadership does not have a vision outside of advertising and its internal culture does not reward long-term maintenance of products.
>> I don't see this culture that HN seems so convinced about.
It started with Google Reader and has become ever more obvious since then.
When Google launches a new product, observers in tech make bets about how long the product will be around before Google kills it.
> Stadia failed as a business because it failed to get adoption by developers and gamers / users.
Nonsense. It failed to attract developers because it was an unsure new platform so of course most of them wouldn't spend hours to develop for it. They were always going to need convincing (with money of course), which took Google too long, but around a year or so in there were multiple heavy hitters in the form of Red Dead Redemption 2, Hitman, Ubisoft's entire catalogue (and an Ubisoft+ integration), EA Games' most new games, Destiny 2, alongside a ton of indie games. By the time Stadia shut down, it didn't really have a catalogue problem. Problem is, Google bungled the rollout and took too long to start doing this - they banked on the initial release being highly successful, but didn't start massively investing in games until later on.
For gamers, fear of Google shutting it down played a part, of course. But so did all the extremely negative coverage, from everywhere, that Stadia is dead on arrival which only reinforced that fear. Had Google actually told everyone their shutdown plan (reimburse everyone for all games), a lot of gamers would have overcome those fears (what was there to lose, really?).
Technically, the platform was amazing. To this day the best UX by far of out cloud gaming platforms. Quality upgrades were lacking though.
To sum up, Google failed in their strategy and marketing. Had they 1) promised to reimburse everyone in case the platform was shut down before X years 2) given away Stadia Premiere packs to anyone they can (like YouTube Premium subscribers, etc.) 3) enticed big studios to port games, all things they eventually ended up doing, Stadia would have been a roaring success and would be undergoing a hardware refresh at the moment. Instead, it just reinforces that Google sucks at b2c and shouldn't be trusted.
>> For gamers, fear of Google shutting it down played a part, of course. But so did all the extremely negative coverage, from everywhere, that Stadia is dead on arrival which only reinforced that fear. Had Google actually told everyone their shutdown plan (reimburse everyone for all games), a lot of gamers would have overcome those fears.
It is a self-reinforcing vicious cycle. Stadia was yet another demonstration of how Google has developed its current reputation.
>> Technically, the platform was amazing. To this day the best UX by far of out cloud gaming platforms.
That's what is so sad. Google's tech is amazingly good, but their user support and reputation are terribly bad. Bad enough that, in many cases, it breaks the business.
That is exactly what it is: a meme. Look at most products / services / hardware on there. A lot of entires are things that have newer version, were renamed or merged with other services / products, or features included in core products.
Some were discontinued, but they didn't have enough use. Only a few user loved products were discontinued: Inbox, Reader, etc.
Yes. It being a meme is exactly the problem. It just gets repeated over and over with no regard to the facts, in a kind of positive feedback loop.
If you did the same for Google's peer companies like Amazon and Microsoft (using the same criteria and same level of obsession), you'd find that the Amazon Abattoir is doing brisk business and the Murdered by Microsoft website would need pagination. But for whatever reason, when those companies launch a product, their past failures are not trotted out. When they inevitably kill their failures, the reaction is "about time, nobody used that product anyway" rather than the mass hysteria.
This is Google Glass! A product 99.99% of HN readers would have claimed was killed a decade ago. "Killed by Google" has been claiming it was dead for a long time, but I'm sure will now dishonestly find a way to double-count it.
I suppose "nobody used that product anyway" is way more true of Microsoft's dropped projects than Google's, and that's important.
Everyone here can name Google products they loved which they can't use anymore. For me, the big ones are Google Reader (RSS), Google Inbox (bundles, and the Travel and Shopping email categories), Google Now (flight information), and Google Hangouts (SMS integration).
For other companies, not so much. I can name minor things, like I miss macOS's old 2D virtual desktop grid. And an AirPort router would have been nice. And I'm sad Amazon Go is being shut down. But nothing on the level of those Google shutdowns.
> "Killed by Google" has been claiming it was dead for a long time, but I'm sure will now dishonestly find a way to double-count it.
I haven't gotten around to it yet, but my plan was to merge Google Glass entries. The holy trinity of Google Glass Explorer Edition, Glass OS, and Google Glass Enterprise Edition will become "Google Glass."
I completely understand the "google kills products" meme.
I just think that "google doesn't reward maintenance" doesn't lead to "google kills products" except in very general examples. For the most part Google is maintaining products, that's most of what most people do.
The reasons for the killing of products are I think much more nuanced and numerous.
Stadia also failed as a technical solution on various of the boring bits. Stadia did not even support the new Chromecast for at least half a year, so you were expected to get the older Chromecast Ultra until then.
I was employed at Google at the time, and within Google this was treated as a matter of fact problem that was not that important. The planning was to support this most likely in the first or second half of the next year. I found this shocking, and to me it made it clear that Stadia was unlikely to succeed if it couldn't even support its main hardware components close to launch.
In comparison, Microsoft lost billions to get into the console market. I don't see the same drive or commitment from Google. If they had the same red ring of death problem the first xbox had, Google would have cut their losses and jumped ship.
Not evaluating whether the meme is true or not (or it’s spectrum truthiness), it isn’t just HN where you will see this meme, it’s also common on memegen as well.
> I don't see this culture that HN seems so convinced about.
Isn't that perhaps because you're completely immersed in it to the point where it's invisible to you? To borrow David Foster Wallace's metaphor, fish don't see water either.
What outsiders (like us at HN) see is the external manifestation of culture. People naturally speculate about what they think is going on inside the chocolate factory, but the external part is very clear - a self-reinforcing vicious cycle where things get killed off leading to mistrust in the permanence of future things, leading to their failure, leading to them getting killed off.
If you discontinue Google Play you discontinue Android so it won't happen. OK, you could extract the app store and discontinue all the other parts that I know exist and I never used, or split them in two, apps vs media. The app store is going to stay until Google will support Android.
This may be shocking but truth doesn't matter, it's about perception maybe users don't see it yet.
The second problem I see is in creative products, if the success bar is 100 million users, how can you take chances on anything, basically before you start your already constraint while small players can grow normally.
How much do you know about the rest of the business though? Even with only 10 to 20 teams, it’s easy to not know specifics about what’s happening in the rest of the business.
That’s problem: there is impact, but it is not recognized well. For promos higher than L4 you need bigger projects. Stringing together a bunch of medium impact bug fixes won’t get it done even if the net effect is very high impact. You only get credit for something you measure and if the quantity is big enough.
> If there is "grind" that is truly not impactful, then why is it being done?
Because it's hard to measure and even harder to predict. E.g. Something that fixes Youtube's scam comment problem where scammers impersonate creators and lure viewers into telegram and whatsapp to scam them.
It won't be working immediately, so someone will have to work on it for a while and iterate against the scammers' countermeasures until it gets too costly for them and most of them give up. Should someone pick it up?
It's not immediately impactful as it won't increase engagement and ads, but it's probably going to prevent a big PR issue when some media company picks up the topic and highlights Google's inaction and their decision to tolerate the scammers and their crimes on their platform.
Fixing it has no direct impact on income, and the problem it mitigates hasn't happened yet, why would anyone work on it? And in the real world we see: they're not working on it (or it's taking them longer than a year, who knows).
I think you might underestimate the number of people working on this stuff. There are rarely trade-offs between, as you suggested, working on ad engagement and YouTube scam comments. Both will have teams working on them.
At a very high level you could argue that the number of people being allocated to each may not be right, but it's not that one or the other doesn't have any resources.
From what I've seen, people at Google are very cognizant of these issues, and the only real reason that the public may see progress come and go is that these are hard problems and the work against malicious content is sadly never-ending.
Thank you for your perspective. It's frustrating from the outside with the lack of progress (or even acknowledgement, really), and it feels like nobody is working on it (or cares). I'm sure they're hard problems, but on the other hand I'm sure Google has lots of smart people, and they have sheer endless resources and own pretty much every part of the user journey and the stack, and the attacks aren't even sophisticated.
Amazon had a similar issue a while back where well-rated seller accounts got hacked (or sold?) and started offering vastly underpriced premium items with text added somewhere to get in contact with the seller, trying to get people off of the platform. But to their credit, they dealt with it pretty swiftly, and I'd consider it a solved problem, I haven't seen it since.
The problem exists at the managerial level. If no VP is going to be promoted for Google glass improvements then they sure as fuck don’t want to be involved. When managers are better off shifting budget to projects that can help them progress, these projects stagnate and ultimately get cancelled.
Maybe, maybe not. Again I'm not thinking so much about the very top of the company as almost everyone is not in that position and never will be.
For me, it's not necessarily true that showing impact on a completely new feature is easier than showing impact in improving an existing one. It's a lot more work to measure new things as we don't have the tools to measure them yet, or as you have to introduce new measurements to colleagues and get buy-in that they are the right thing to be measuring. On the other hand, saying that I reduced an existing error rate by half is very easy to communicate, understand, prioritise, and even reward.
The problem honestly has nothing to do with SWE/PM incentives. The problem here, as noted, isn't going from 0->1. The problem is business strategy and leadership, and the ability for multiple stakeholders to rally around a xfn initiative to ensure it receives dedicated care & feeding to take it from 1->100.
When you step outside the eng bubble, the incentives problems absolutely apply when discussing issues Google has creating mature products.
Another issue is about the clients. In the AR-space, "Enterprise customers" is the secret code to say "military clients".
If Google has the policy to not work with the DoD, then it makes its options very limited (Magic Leap doesn't have such restrictions as far as I know).
The same way they likely didn't appreciate when they were associated with Boston Dynamics.
So, if you can't sell your product for its main useful usage, then it's a bit difficult :/
Google has no such policy. Google Workspaces has active DOD contracts (and DOD is featured as a case study on their site) and active pursuits in other spaces, employee concerns very much notwithstanding.
Sure you can get promoted for launching a new thing, but you can also get promoted for fixing a first pass implementation, for scaling something from a small user base to a large user base, and so on, as long as you can show the "impact" of the change – whether that's revenue, savings, productivity, improved UX, etc.
It's possible that up at the top there's more of a focus on shipping new things, but I don't think that's something that 99% of Googlers need to obsess over, and it doesn't reflect the bulk of the work actually being done.