It's hard to take this post seriously when basic facts are off.
> To support this company strategy, Google makes it really, really easy for anyone to start up a new project. The culture of “20% time” gives people permission to work on ideas outside of their primary role. There’s only the lightest of reviews needed to launch a “1% experiment.”
No, it is not super easy to start a new project. I'm in Google Cloud and we're pretty focused on executing team priorities. I can suggest new work-streams and write arguments in their favor, sure. But I can't just randomly add a feature or go away to work on a new product for a while. Every public thing in Cloud, at least, is carefully reviewed with a large checklist covering privacy, security, accessibility, API style, integration with existing systems like metrics, legal, etc. The simplest cloud service can take an extra six months or a year to get past these blocks. These requirements also apply to features. What we get for this is quality and consistency between products/features.
And 20% time isn't really a thing any more. You can get permission from your manager to work on something else, perhaps. But more likely this sort of thing is more like 120%. It's not at all as lackadaisical as it's presented to be. The most I see around my team is people volunteering for large-scale clean-ups. But these are more slack-time projects like 5% projects. It's not totally gone however. I think the person who made Google Maps zoom out to a globe did it as 20% project, and I believe adding CitiBike support was similar as well.
We are also super careful about experiments. Every Google Search experiment has to go through a senior management review. In my team, every flag rollout is discussed and approved before rollout.
Insofar as these tidbits are presented as evidence for loose leadership, they don't make sense. The ship is fairly tight, at least in Cloud.
[Bias disclaimer: I have worked at both Microsoft and Google]
Google Cloud is different from Google. Cloud is where Google is learning the hard way that being Googley doesn't work when your customers are businesses. This whole article struck a deep chord from me - in fact, I've given the exact same answer to countless interview candidates when asked about companies I've worked at that Google gives ICs far more leeway to decide what they build compared to other companies who are very top-down.
Google culture, in my personal opinion, struggles with enterprise sales and the customer support, long-term stability, and overall business plans it takes to attract enterprise users, convince them to use your product, and convince them that you're serious about supporting the things they care about. This is quite obvious when you compare them to a Microsoft or an Oracle. Cloud is the first major business group that has had to come to terms with this, and they've taken several steps to address this. Some people have complained that the culture in Google cloud isn't the same as in Google at large and that it feels more like certain other companies, but what they miss is that this is absolutely essential, and if you tried to run a cloud compute product with the traditional Google model it would (in my opinion) fail miserably.
"being Googley doesn't work when your customers are businesses"
Cannot upvote hard enough. I have had to work with several Google PMs and SWEs on large contracts for different companies. They all talk down to you, think you're confused, and if you ask for a feature to reduce costs it's always: "that'll take 6 months, but here's the solution we like and you should try it and it's really cheap (and no we definitely don't want to lock you in of course)".
Being Googley is a lot about Googlers satisfying their own curiosities and performing well against their own metrics that they define. Sometimes that's correlated with what the customers actually need and want. Other times, it's really not, and as a customer you should fire your Googler. Don't worry they'll still find a way to keep themselves entertained.
I recently did interview loops at both Azure and GCP and the difference between the people were astounding.
While Google takes way too long to make a hiring decision, all of the people I talked to seemed intelligent and very thoughtful. I did a Microsoft 'hiring event' and holy hell it was a clusterfuck. I got ghosted by one manager, another was extremely disinterested and seemed like they were interviewing me to check a box.
Another asked an algorithmic coding question they themselves did not understand, I had to very politely argue with them, I ended up having to show them step by step how my solution was correct (much to their surprised amazement). The system design interview was done by someone who clearly doesn't understand system design.
I cannot emphasize enough how much their engineering leads came across like a joke compared to other companies. Maybe I just got really unlucky and there are good people who work there, but I'm a bit frightened by what I saw.
Worked at Google Cloud a while ago and GitHub (Codespaces, which is built on Azure), I couldn't agree more. Google Cloud is so technically and operationally superior, and the average engineer quality was _much_ higher. The obvious downside is that because the bar for GA is so high (because everyone is, or is trying to be, clever), everything lives in beta forever and you can't ship anything on any reasonable scheduke.
Google and Azure are both just light-years behind AWS , having worked on and with all three platforms and dealt with engineering staff all the way through.
Google doesn't really build things for other engineers, hiding important details of system operation. Azure resources just don't spin up.
AWS docs are a byzantine nightmare sometimes, but everything that's in the documentation works exactly as it says it does, and it doesn't go out of its way to go "look over there" when a mode of operation has downstream implications.
From what I've seen of Azure as a customer, the quality sucks and it's impossible to convince them to plug gaps no matter how trivial to fix and how obvious. Meanwhile I've had at least one ticket become a permanent fix in AWS almost without asking.
However...
I just switched back to AWS and it's like stepping back in time from Azure.
The AWS Console structure, layout, and organisation is their internal corporate structure.
The Azure Portal is designed to reflect your corporate structure, not Microsoft's.
With AWS if I have some VMs, associated S3 buckets, and an RDS database, I have to click around a bunch of different web portals to find them all! Unrelated resources are all blended to together into flat lists with gibberish internal system-generated names.
It's like clicking through the tables of a SQL database instead of using the actual client application.
In Azure, everything is nicely organised into Resource Groups. The VMs, storage, load balancers, databases, etc... all together. Everything is shown with display names instead of IDs, and all their individual consoles are (mostly) consistent. Logs, RBAC, etc... are generally all in the same place and work the same way.
>I did a Microsoft 'hiring event' and holy hell it was a clusterfuck.
Microsoft's recruitment process is notoriously woeful. I wouldn't have believed half the things I've heard only it happened to a friend of mine. The interviewers never showed up to the interview, and terminated his job application when he told them the rescheduled date for the interview wasn't suitable.
That’s an interesting angle, how ones B2B dedication and competencies align with business customers who have a lot on the line vs flippant and irritated end consumers playing with free products. The stakes are definitely not equal there.
In case anyone from the Photos team sees this: Photos is the reason I still use a personal Google account, and I pay for Google One. It just works. Everytime I notice a change, it seems like it tangibly benefits me. When I don't notice a change, everything is smooth, fast, and easy.
What Google also misses in terms of B2B is the non-technical side of keeping your clients close (or, more to the point in Google's case, to get said clients in the first place).
Google somehow still thinks that business clients will flock to them just because they're called Google. That's not how it works, you have to keep close personal relationships with said business clients. If that means having a heavy-alcohol infused night out in some expensive resort with the CTOs (or the CTO-equivalents) of some of those potential clients then that's the way do it. To say nothing of the government clients, who come with their own idiosyncrasies.
> ...if you tried to run a cloud compute product with the traditional Google model it would (in my opinion) fail miserably.
Google does sell G Workspace to enterprises? YouTube also seems to do well with enterprise-y stuff involving the Movie and Music industry, as another example.
GCP is all-enterprise zero-consumer, sure. But: GCP's got execs from SAP and Oracle at helm, and I believe, it is a problem that, over the shorter-term, they can throw money at (by hiring execs from Dell / Oracle / Cisco / HP / IBM / Qualcomm, as at that scale of spend, enterprise sales is mostly a game of who knows who).
Back the day Google salesman would drop in like and say “hey, we’re from Google, perhaps you’ve heard of us”. Seemed like their primary gig was apologizing for the death of the Google Search Appliance. They weren’t very empowered - on one project that I worked on, the sales-dude had so little juice that the SE put stuff on his credit card to do a demo.
Now, they seem to have a solid enterprise sales team. The only weak point is they rely a lot on the channel partners. Usually that’s fine, but I prefer one throat to choke.
> Cloud is where Google is learning the hard way that being Googley doesn't work when your customers are businesses.
I can't speak for Google, or about what Googley is. But I have worked a decade in the Danish public sector, and now for around a year in European Green Energy, and I think there is just a massive difference between Google and Amazon in this regard.
Denmark is one of the most digitized countries in the world, and I was there for the transition from self-hosted iron in the basement, to renting space for your iron in dedicated server farms, to going full Azure. Azure was/is the logical way to go for a lot of European Enterprise organisations because it hooks so nicely into the Microsoft infrastructure you undoubtedly already have. There is no real alternative to Microsoft Office in Denmark, Google Docs could (maybe should) have been, but I'll get to that later, and once you're buying the Office365 package as well as Windows licenses, you're going to have massive savings by going with Azure rather than AWS. This is partly because we don't buy licenses directly from Microsoft, but through partners and because Microsoft has some of the best B2B support in the world. I know Microsoft gets a lot of hate here on HN, and maybe rightly so, but it's by far the best tech company I've ever dealt with from a non-tech enterprise perspective. When team was rolled out, it was turned on for users automatically, which is a big no-no in Enterprise IT, but within an hour of it happening Microsoft had given us a custom hack to turn it off in our setup. Not long after, probably because the entire world of Enterprise was on the phone with Microsoft, it was off by default. But compare that sort of support level to Google, where you can talk with an automated form... Google will lose so hard that it's not even considered an option.
Amazon and Google were interesting in this regard. Because the where Google didn't really do anything, Amazon quickly realized that they couldn't sell AWS to Enterprise organisations if they didn't compete with Microsoft on support. Because that's sort of what you sell to the IT department in Enterprise Organisations. One thing is the product, the other thing is the ability for the CTO to tell the organisation that they are on the phone with Microsoft/Amazon headquarters and getting hourly updates when something goes wrong. That's of massive value, and Google doesn't do this. Amazon similarly didn't comply with the GDPR as fast as Microsoft. At first it seemed like Amazon didn't really think it mattered to them, and that we would still pick their cloud products, but it didn't last very long, and within a year, they became more compliant than Microsoft. I remember it well, because we had a few things in AWS at the time, and we needed to help filling out some "red-tape" and Amazons reply was basically "kindly fuck off", and then out of the blue 3 months later, we suddenly had a Danish AWS representative and free legal counseling from one of our major EU focused lawfirms contact us to help us get it done right.
Google didn't do anything like that. They sort of try in Education, but we're currently well into a big chromebook + Google Education and school start "scandal" of sorts here in Denmark. Because one of our cities failed to pass the national GDPR requirements and were basically forbidden from using Google Education at the start of the school year back in july (the school year starts in august). This didn't exactly happen, because it turned out they couldn't start the school year and that parents care more about education than whether or not Google snoop on the schoolwork or not. It was also sort of silly because it was just one municipality out of 98 that were forbidden from using Google Education and chromebooks and there are a lot of other municipalities that also use it. Anyway, that is sort of unrelated to what I want to use the example for, because the interesting part to me is how Google hasn't really solved the issue.
Google Education and Chromebooks are a massive business by Danish standards. The Public Sector is the biggest spender on IT, and if you can secure just half of the schools in the country for Google Education, that's one of the biggest opportunities for IT sales you can get here in Denmark. And Google hasn't bothered solved this issue? I think the very likely outcome is that all the schools are going to switch to Microsoft and Office365 and run Linux on their Chromebooks in the coming years because of this. I think this because Google isn't the first company to fumble the ball on Education. Apple did it first, and where almost every school used Apple products a decade ago, now they use none. Apple's problems weren't GDPR problems, the GDPR didn't exist back then but we did have national privacy laws and they were in perfect compliance with those. Apples issue was instead that their products didn't integrate into Enterprise IT policies, and since they didn't want to offer support or change ways, they were slowly rolled out and replaced by companies that did.
I think this may have been fine for Apple. They don't seem to be big on B2B sales, but for Google it's just so strange.
>> its hard to take this post seriously when basic facts are off.
For context I don't work and have never worked at Google. So I can't speak to whether 20% time is a thing or not, or whether managers manage or not.
The article does however reinforce a fact which is obvious to pretty much everyone outside Google - which is that Google has a perception problem - in my perception "everything gets cancelled" [everything except the super successful money machines that are Search and YouTube. Gmail has enough benefit to search, maps as well, but outside that what has survived?
GCP suffers from this perception (and the perception of lack of phone support.) given the lackluster performance of GCP its hard to be _sure_ it'll survive and that's a problem.
Annecdotally i'll never use GCP - you can thank Reader et al for that. To be fair I also would never use it because AWS and Azure have better support.
I understand why you want to correct the record, and that's fair enough, but the details in the article are irrelevant. Regardless of the "how", the "what" remains the same,which is that I have zero trust in Google to act in _my_ best interest, and that means never using them for anything important.
Google would have to be more than 25% cheaper than Azure or AWS, and I don't mean "at least", I mean more than, as in closer to 33% or ideally 50% cheaper... before I would even contemplate them over the competition.
> To be fair I also would never use it because AWS and Azure have better support.
This is the salted earth that GCP have to till... Google sucks miserably at customer support and anyone that has interacted with them knows it, from the millions of android users who would flinch at the mere sight of a terminal prompt to the developers that have wrestled with GRPC services and who keep giving their code a chance. We all go through life see the facets glinting in the sunlight, weird error messages, a bad two factor interaction, terrible search results, business destroying story #12 this month from GCP + <some other service at google> ... little reminders of "oh yeah thats just Google, sucking as usual" and continue to remember how bad they are at customer support.
The very earth itself has been salted with the tears of the millions who have had sad interactions with Google...
And this is the ground GCP is trying to grow a business in...
Google suffers from a perception problem for consumer-oriented products, agreed. I don't think anyone faults Google for haphazardly closing down enterprise products. The enterprise world runs differently - deprecations have long timelines, are well-telegraphed ahead of time, and sudden deprecations may come with contractual penalties. Which Cloud products are analogous to Google Reader?
Stadia, I think, is the biggest example of a Google product failing bc of this perception. But it had other, big problems as well.
>> I don't think anyone faults Google for haphazardly closing down enterprise products.
I think you meant "consumer products" here? I mean, I'd definitely fault them for haphazardly closing down enterprise products.
I confess I don't differentiate Google into "Google Consumer" and "Google Enterprise". So, to me, an issue with say Gmail or Reader is a "Google issue" - not a "Google Consumer" issue. I don't think to myself "oh yeah, well Google+ is a consumer product so closing that has no bearing on their enterprise offerings".
So sure, GCP maybe an entirely different beast. But my perceptions are for all of Google. If you want GCP to be judged on "non-google" standards, then perhaps don't put Google in the name...
To be clear I've never tried GCP. It could be 10% of the price, run 10 times as fast, and have magnificent support for all I know. However I likely still wouldn't use it (for anything meaningful) because I don't _trust_ Google. This is a trust issue, not a price or performance issue.
> The enterprise world runs differently - deprecations have long timelines, are well-telegraphed ahead of time, and sudden deprecations may come with contractual penalties.
Is Chrome an enterprise product?
Because the whole Manifest v2 deprecation timeline was a debacle, considering many businesses run critical extensions. Originally Jan 2023 (without GP override), now June 2023.
Even when I was there a couple years ago, 20% time was already basically dead. They had to be approved by those above one's direct managers, needed compelling justification, and getting the go-ahead was pretty rare. This was even for things not in the main products. In other words, 20% projects are the opposite of what they started out as.
I also saw Google shutting down or downsizing a lot of small projects that weren't related to the core business (ads, search, cloud), though that might vary per PA so my read might be off.
When I worked at Google in 2007, it was a known fact among employees that 20% was not a real thing. It was 120% time even then. You could in theory work on your own thing, but you basically had to do it on your own time.
And I did actually do the 120% time thing for a few months because there was a specific project I really wanted to do, and it would have been nice fit for Google, but my attempts to stir up interest within the company and get any sort of resources were basically dead ends. One of the reasons I eventually left to join a startup.
20% time never made much sense to me as a concept. If the 20% is not part of your performance evaluation, then that would imply everyone is on 80% time. What does that mean and how is it worked into the process? Would you get a 125% score if you did only your normal work instead of a side-project? If you failed to do 20% of your work, do you get a pass?
It seems to me that your work should get graded on some standard, and what % of your time you spent on side-projects should be irrelevant.
In theory the idea was that everybody should want to do a 20% time for the good of the company, and thus failure to seize the 20% opportunity would count against you in the promotion process. In practice though your promotion was really just graded against your main job and nobody cared about 20% time, thus actually doing something in 20% time would be bad because it would hurt your performance for your main job. Even in 2007 everybody would talk about how things were so much better in 2004 and if only you'd seen it then, and so I imagine that maybe back in 2004 you could do 20% time and it would be beneficial for your career, but you'd have to ask somebody who worked there at that time.
As far as I could tell there were a bunch of very talented, very senior engineer who could basically just do whatever they wanted as 100% time, and they did a lot of really great stuff that was helpful to the company as a whole but didn't fit neatly into a specific team. But it didn't seem like that was accessible for mere mortals, nor would you be able to make as large an impact if you were only working on it one day a week.
>Google Cloud is not the real Google. Culture-wise, you are closer to Oracle.
Incorrect. Google's culture has shifted. Culture-wise, Google is closer to Oracle than what people think is Google.
20% hasn't been a thing in Google for a long time. Exists only on paper.
Source: I worked at Google when we all got the memo that all information access within the company is now on a need-to-know basis, and that accessing information that you don't need to know for your job is a fireable offense.
Default level of sharing technical documentaiton is fully private, so tech docs are not searchable in company search engine (read this again).
Good luck re-using the code without design docs. Also good luck starting a project if your team doesn't own the data.
Hot take: Google doesn't have a strategy, period. Neither copmany, nor product.
To further your point — 20% time was barely a thing in 2010… And they killed all of the OC/Founders awards subsequently so there was really no focus on maverick innovation from then on. Coincided with the end to oysters on Fridays
Google cloud has a strategy / OKR process. It’s called Amazon Re:invent.
The blog had some good points but was totally off of Cloud being bottom up. It couldn’t be as it was so far behind others — you don’t have time to think up something new when the competition has already done something. Decisions came from the top down.
The author was right in that there’s a lot of smart people there. By far the best place I’ve been for that.
This shouldn't be surprising. Perhaps comparing it to Oracle might not be quite right, I leave that to people who know, but Google is pretty obviously not the startup/disrupter of 2000-2004 any more. It is now 20+ years old and huge, and that means it is going to be somewhat like any very large organization: internal politics/empire defending, standardised compensation and management, mission, vision and values, etc. Your experience likely varies greatly depending on where in the org you land and who with - some people will come out writing blog posts and some people will find their happy place and never leave.
The interesting part around Google is that they had a kind of mythical status amongst techies that's lasted quite a lot longer than the early days. But they're just another big company now.
Amazon is also slowly moving to this "private tech document" as a side effect of using Quip instead of putting them in the company wiki searchable for all. I hate this.
It's sad how once a great company decided to shift towards mushroom management and coding sweatshops like Amazon in just a few years... The amazing Google of 2008 is no more.
Exactly, hence no true Scotsman: "Oh, it's not a true Google org!".
No, it is a Google org. Whatever was said in the top comment was applicable to Geo a few years back.
And the infamous "need-to-know" memo, which required everyone to start siloing tech docs and made using the internal search engine potentially a fireable offense (!) was company-wide.
Having worked at both, 20% time is the only part of the article that is somewhat off-base.
In particular the following string of paragraphs match exactly what I’ve seen at both:
> With all that up-front planning, the actual product decisions and working code were reviewed much less. Only the other PMs on my team reviewed my specs, and only me and the tester (also new grads usually) reviewed the working changes before they were added to the branch. For what it’s worth, I think this is why the quality of the details on Microsoft products is often lacking—a random PM made a decision and no one bothered to push back on it.
> Contrast that with Google.
> At Google, I never saw a strategy document. I went to every Friday all-hands, but I didn’t hear the leaders talk about a broad vision of what we needed to build. The people above me didn’t set direction and ask me to follow it.
> Instead, leaders encouraged teams to generate their own ideas. The founders had ideas of their top priorities and worked with the relevant teams on those, but the list of the top company OKRs was a subset of all the OKRs, not a roll-up. Some people were not working on anything the company cared about strategically!
Again, I can only speak for Cloud. I find product changes are reviewed by a dozen people at least and take months to finalize. It can be slow, but usually we're sure why we're making the decisions, and where risks lie. A large weight is given to customers asking for features, so they're in the loop.
It's absurd that there's no Google strategy documents. There's company-wide and division-wide OKRs that get major events to publicize them, with large Q&As and all. It's easy to learn what the priorities are and why. But, usually you just pay attention to what your local group is working on. At a smaller level, every feature has a business rationale backed by customers requesting them or hard dollar figures.
I really don't know what's going on here. Do people think Google runs the same way as it did in its early years? It's a fairly mature company with a good track record of revenue growth, especially for such a massive company. These things don't happen haphazardly.
We seem to be looking at the same text and drawing wildly different conclusions from it.
When the blog post says "the actual product decisions and working code were reviewed much less. Only the other PMs on my team reviewed my specs" referring to Microsoft, to me that lines up with you saying "I find product changes are reviewed by a dozen people at least and take months to finalize." referring to Google. The distinction the author is drawing is that Microsoft doesn't review day-to-day engineering work like this, because fundamental product decisions are made at a higher level, while smaller design decisions might be made by individual PMs or engineers and make it into the product without ever being reviewed.
When the author says "The founders had ideas of their top priorities and worked with the relevant teams on those, but the list of the top company OKRs was a subset of all the OKRs, not a roll-up. Some people were not working on anything the company cared about strategically!" referring to Google, that sounds very aligned with you saying "There's company-wide and division-wide OKRs that get major events to publicize them, with large Q&As and all. It's easy to learn what the priorities are and why. But, usually you just pay attention to what your local group is working on."
+1 to basic facts being off. The line that stuck out to me was:
"Some APMs had convinced company leadership that we needed a Facebook competitor."
Ummmm... no. This was very much a top-down strategy choice from Larry on down. There's no way the product goals COMPANY-WIDE is driven by anything but an executive edict.
so google still does that 20% thing? I thought that went away when google's unofficial company motto stopped being "do no evil"... in the time before Schmidt CEO.
Google has cancelled a lot of things. And they do kind of suck at product strategy. The worst part is, they don't even realise when they have something great that's on the verge of gaining traction (Hangouts and Stadia IMO could have both become enormous with just a little more persistence and marketing).
That being said, on the technology front, they're still better than anyone else. Chromecast just works in a way nothing else does to get media onto your TV. Google is better at syncing things in the cloud (Google Drive/Docs) than anyone else. Office365 always has weird issues with synchronising documents in the cloud and is a huge pain in the ass whereas Docs works perfectly in real time. Google search is obviously the best. Google Maps is the best. Gmail. YouTube. Android is an amazing OS even if some OEMs add crap to it that detracts. Chrome is by far the best browser. And so on. I put up with Google's subpar product 'strategy' because it still makes life easier than dealing with MS or Apple's nonsense.
The reason this is true is simply because it's dominant, so every site builds for compatibility for Chrome. This is similar to how Windows is the best OS for gaming. Not because Windows has any technical advantage, but it's simply what developers build for. Firefox would "be the best browser" if it were dominant.
> The reason this is true is simply because it's dominant
And I remember a time where it wasn't dominant, where people went out of their way to install it, even when their OS would shove other browsers in their face...
It didn't become dominant by being a bad browser...
> where people went out of their way to install it
You mean, when Google went out of his way to make people install it. It is delusional to believe that the average PC user installed Chrome for its technical merit. They installed Chrome because there were advertisements everywhere, from subway to Youtube, and, above all IMHO, because ‶the Internet (understand google.com) said I have to install Chrome to go fast″. And that's not counting the heaps of installs that must have come from Chrome being bundled in adware installers and sneakily setting itself as the default browser.
It didn't hurt that yes, they advertised it, but I don't think it's fair to discount the effect that the "technorati" had on increasing its adoption.
That is, pretty much every tech person I know switched to Chrome because it was just so much better than what existed at the time, primarily its faster JS speed with v8, and I think most importantly for actual end-user experience, per-process tab isolation so that one bad tab wouldn't crash the whole browser. These were real, important innovations that had the tech elite quickly adopting Chrome.
So what happens once it's adopted by the tech elite? First, they push for making it the default install on new computers, which has a huge effect. Next, every time Mom or Dad or Grandma asks kid/grandkid for tech support, the first thing they do is say "Install Chrome first, that should help with a lot of your problems" - I certainly know I did it with my family members.
So while I don't want to discount advertising, I also like to say that it's easy to advertise great products - you just need to let people know the product exists, because eventually the product's benefits are self-apparent.
I don't buy this. Advertising is effective but advertising alone wouldn't have been enough to build this much momentum. This is all academic, but I don't consider most people idiots. I think most people installed Chrome because it had features they wanted or a technical person in their life recommended it. I was one of those technical people recommending Chrome.
> I think most people installed Chrome because it had features they wanted
Just go ask people from you family using Chrome why they do so, I bet at least half of them are not even aware they are using Chrome and not just ‶going on Google″.
Chrome is definitely a good browser. It helped evolve the web. But, I don't think that Chrome today is much ahead of other browsers aside from compatibility.
When Chrome initially appeared, it was pretty much on par with Firefox, not significantly better. What made it succeed is great marketing and being forced and pushed down every possible Google channel there was, including being preinstalled on Android phones.
For significant periods of time in the mid-late 00-s to mid 10s, this just isn't true. Firefox's plugin, process and memory management was significantly worse than Chrome until relatively recently. As a result, for memory limited machines (all of them - remember, 4GB was considered huge), Chrome was the only real choice. I personally gave up on FF for 10 years from 2011 or so because I was sick of the plugin manager crashing and taking the machine with it.
Maybe in the middle part of that period. I'm saying that when it first appeared, it wasn't clearly superior. I'd argue it's also not clearly superior anymore.
To bring my own experience to it, I stuck with Firefox all throughout that period and never had significant bugs, slowdowns or crashes. Though I will admit that there was a period when Chrome felt noticeably smoother.
Eh, I think there are other benefits too: excellent extension ecosystem, simple and clean interface, default search engine is Google, generally fast and stable.
Chrome is the iPhone of the browser world - there are probably things in other browsers that work better, or that you could customize to fit your needs, but instead of spending the time to optimize those you can just choose chrome and you'd be pretty satisfied overall.
A lot of other companies cancel things, but until Stadia was killed Google did cancelations that would shock and anger consumers. Why?
1. They don’t tell people that it’s an experiment like Amazon.
2. The experiments feel short. This makes people anxious and hesitant about buying into a Google experiment
3. For paid products, they don’t offer refunds like Amazon. This makes people feel annoyed.
4. In some cases like with Nest Secure, Google had a fire sale BEFORE they made a cancellation announcement. This makes people angry.
The end result is that consumers now feel that it is a risk to buy products from Google. It’s great that Google is finally acknowledging this problem with the Stadia cancellation. Many people were pleasantly surprised that they were getting refunds on the hardware, though I’m sure anyone who bought a game outright instead of just subscribing isn’t too happy right now
Going on a tangent, Chromecast is one of the worse streamer boxes because for whatever strange reason, Google does not realize that multiple people will use the same streaming box and each of these people will want personalized recommendations. Does it have an “add another account” feature? Sure, but unlike every other streaming box it doesn’t just work. You have to make multiple attempts just to add another Google account.
> Many people were pleasantly surprised that they were getting refunds on the hardware, though I’m sure anyone who bought a game outright instead of just subscribing isn’t too happy right now
I believe they refunded software purchases and in-app purchases too. The only thing that didn't get refunded is Stadia Pro.
Gmail is not the best anymore. It barely seems to be worked on.
Google Search is constantly the subject of complaints about terrible quality these days.
Chromecast is pretty great, but it's literally not possible to build a competitor.
Why is it not possible to build a competitor to Chromecast? Sure, you'd need to manufacture and sell a hardware dongle*, but that's not impossible. On the screencasting front, it's possible to screen-grab from Android and send that so you could mimic built-in functionality (not sure about iOS, but if Chromecast can do it then other apps can do it). Casting from media apps would require tie-ins with media companies, which again is tricky, but they're generally a bit pissed with Google so would likely welcome a competitor to the space. You'd have no chance of getting Youtube on board, but then you've got a decent case for anti-competitive behaviour, which EU and US systems are circling Google for anyway.
I'm not saying it would be easy, but it's certainly not impossible, with the right business model and business and political environment.
* It might be possible as a smart-TV app, though smart TVs tend to suck which is why Chromecast still sells units
The biggest problem is their reputation precedes them. Nobody thought Stadia would last so nobody joined the ecosystem. To be fair it was a ridiculous ecosystem where you couldn't port in any of the games one already owned but how many people took one look at it and said "why do I want to buy games on a service that'll probably be dead in 18 months?"
I love the service but it has a fatal flaw (by design): no ability to synchronise my photos to my computer. Given the horror stories of Google accounts being locked for any and no reason at all, with no recourse, I've become extremely distrustful of cloud storage. I use it only with regular or synchronised backup. Google used to offer the ability to synchronise pictures to one's PC, but they removed that functionality. Presumably because more friction in this area keeps customers stickier. They're less likely to leave.
The only option is a full download of all pictures, and they do their damnedest to ensure it's uncomfortable. Albums aren't properly labelled and linked. Labels are a mess. Picture naming is broken. Exif data randomly stripped or altered. Duplicates galore. This process ensures that, even if one is diligent and ensures they regularly download the entire database, they'll end up with a completely unmanageable picture library with dozens of duplicates. And most people aren't that studious.
Have you managed to find an alternative? Your concerns resonate with me but I still don't know the best way to automatically backup all the photos I take on my phone, search them later, etc
I ended up paying for iCloud and buying a Mac. There is an option to automatically sync full copies to Mac. They charge a premium but it works flawlessly.
Caveat 1: it's technically a sync, not a backup. The risk remains that Apple deletes all my pictures remotely, and this deletion propagates to delete my local copies. There is still a recycle bin, but perhaps Apple has some way of deleting even that remotely. I don't think so, but it's a risk. To mitigate this, I periodically download all pictures to "cold" storage. It's fragmented and poorly categorised but at least the pictures are there in the case of a catastrophe.
I worked with several Azure product groups in the past. Although we had a top down roadmap, we were also heavily influenced and guided by customer needs. A big component in our backlog prioritization was whether we had customers who would use, test and provide feedback on private previews released to their subscriptions. With review, SWEs and Product Owners also had enough agency to add functionality/features to the backlog.
Perhaps the Office org is run differently to Azure.
While I feel like this article is a good assessment of Google's culture, I think it's a bit weird that it ignores the prime reason Google shuts down so many products: if it ain't tied to search/ads, it's hard to argue it will make any significant difference in Google's revenue.
That is, Ad Words and their other search/ad products are so phenomenally successful, basically any other product that is even slightly "niche" will be in the .00...001% impact on revenue. That's why if you look at Google's most successful products besides straight-up search (like GMail, Chrome, Android, YouTube, Maps, etc.) they are somehow tied to viewing more ads that Google serves up. So many of their other products, probably Stadia being the best most recent example, may have great technical execution, but they are just afterthoughts and would probably always be afterthoughts, revenue-wise, compared to Google's ad revenue.
I think the one area that is interesting is Google's enterprise products, particularly GCP and Google Workspace. As others have commented, Google has great technology in these areas but they often just don't understand how to work with businesses in a way that businesses expect.
With regard to Microsoft vs Google planning, I wonder if a lot of that difference comes down to their rolling release of the web vs the previous boxed version releases.
I remember in the 1990’s a release of Windows or Office was a big deal. There would be a ton of press. The boxed version would be at all the computer stores (and even at stores like Walmart). There would be tons of reviews talking about if it was worth upgrading, etc. With all those things, you want your release to go off perfectly. You would like to avoid half-baked/buggy features, and also have enough features to wow the reviewers. A bad release would be a decent hit on revenue and market position.
Now compare that with web releases. You can slowly roll out features when they are ready. You can fairly easily A/B test then with a small percentage of queries to make sure everything works ok. Most people likely don’t even think about versions of Google web apps. In those circumstance me, allowing more freedom for people to work on new ideas with a minimum of overhead is likely a win.
I've come to realize that the damage due to Teams to the Microsoft brand name is enormous and in no way recompensed by the market share that it captured during COVID.
Can you expand on that my org switched to Teams during the Pandemic and it has basically transformed my workflow. I work as an Engineer (The non software type) so maybe my workflow might be different form Software people but Teams has basically managed to almost completely displace email, phone calls, sharepoint and a lot of in person meetings.
Things like sharing documents between people working on a project together. We used to use combination of email, one drive and sharepoint document libraries.
Teams streamlined this immensely with it's built in groups it is really easy to create a new group for each project and it stores all the documents, chat messages, meetings etc all in one place and it's all archived and searchable.
Even stuff like screen sharing and screen recording (so people can catch up on meetings they missed) work really well - we used to have to use Skype for bussiness to do this and it was not great at it.
I'm not a Microsoft fan but they really nailed Teams in my opinion I'm surprised to hear people are negative about it.
Teams may work well if you are fully in the Microsoft ecosystem and have a dedicated enterprise level IT organization. That is not the kind of place I work.
Sometimes I have to use Teams for meetings and it is incredibly frustrating. First, it makes me sign in to an account. Then, it is not sure which account to use since we have a corp account and I did the sin of buying office365 with my work email. Eventually I reset one or more passwords via its login flow and it lets me in (just using the password does not work). On the call, background blur is not usable from the menus or UI, but a StackOverflow post saved me: the keyboard shortcut still enables it. Also, Teams insists on always starting at startup and I can’t make it go away.
Fair enough, you're right I'm in an Enterprise type organization with dedicated IT. Teams starts up logged in when I boot up my laptop (I sign into it on my phone via my work email address). It "just works" for me but IT has probably done a lot of black magic in the background to make it happen.
Azure ML studio which is another Microsoft product I use on rare occasions I have issues with there is some type of 2FA setting my org has enabled and usually takes multiple attempts (text messages) before it authorizes me. I frequently have to reauthenticate during the day which also annoys me.
A lot of folks on here don't live in a Microsoft-first world (I'll include myself in this). Teams really isn't enabling anything we weren't already doing pretty smoothly with some combination of Slack, Zoom, Google Meet, etc.
Add onto that a few legitimate bugs/design flaws, a natural preference for the familiar solution, and just a sprinkle of anti-Microsoft bias and you have a recipe for the perception that Teams is awful.
Obviously some part of their strategy is working considering how much usage they get of such a poor product. Building a good product just doesn't seem to be part of the strategy.
The author of this blog has been the head of PM at Asana for nearly 9 years, during which they've completely failed to come up with a sustainable product strategy and have been bleeding money for years.
"... the world would be a better place if people spend less time on failed projects and more time working on things that make a difference."
Asana was founded in 2008. It seems destined to fail, and will have largely wasted a large number of people's time, and seems very unlikely to be a "thing that ma[de] a difference".
"Asana lost an incredible $285m in 2021, $210m in 2020, and $118m in 2019. They're on track to losing even more with over $370m in losses booked for the trailing twelve months. That's closing in on a billion dollars in losses over the last four years. Madness."
There’s a joke in the first episode of Silicon Valley where some mundane startup oversells their mission as something about enabling humanity to thrive. I saw an interview with Mike Judge where he said it was just a copy of an actual experience he had where a founder was telling him all these grand visions about how he was going to lift up humanity and Mike thought it was some sort of charity. But when he asked again what the company was, it was a todo app! Mike didn’t name the company, but I Googled the mission statement he mentioned and it was clearly Asana.
Not to mention asana just isn’t amazing either. It’s such an elementary pm tool given how long it’s been around and how much money they’ve spent, it really makes you wonder where it’s all gone.
At my last startup we tested a number of options and decided to go with ClickUp. I personally preferred the philosophy of Linear, but ClickUp can do way more and it’s quite good. Their feature velocity is really impressive too. And we did run into bugs and performance issues, but support was responsive and helpful.
Honestly if you get a very skilled atlassian operator you can do massive scale project management pretty easily. Everyone loves to hate on jira, but it’s crazy powerful. It’s just easy for it to go to shit unless you have someone who is both highly competent and highly opinionated running it all.
> You can count on your two hands how many small to mid size SaaS/tech companies actually make money.
I think you’ll see this change slowly in the next 2-3 years.
I’ve been running a small/medium company for 7 years now, bootstrapped but with a small $1 million seed round in 2015.
During the 2020/2021 bubble, our seed investors were very angry at us for being profitable. (Angry isn’t an exaggeration)
After the bubble bursts, those same investors are now praising us for being so “cash efficient” and profitable.
Founders are heavily influenced by the recommendations of advisors around them. Now that the entire VC community is condoning/punishing high cash burn companies, you’ll start to see more conservative ways of operating a business become more common, since that’s now what VC, PE, etc all want to see right now (and for the foreseeable future, at least until interest rates go way down)
If you grow merely sustainably and profitable, many VC won't fund you, as track record doesn't show the curve they need to offset all the fails and/or market you to the underwriters for IPO.
This headline made me think of Google as one of those money laundering apparatuses you watch mob / cartel movies about briefly, they overspend on “projects” and make it just functional enough to look legit. Course if I were a true evil money launderer I almost want to say Amazon as a “shipping company” of sorts fits the bill too.
Google has a company strategy, rather than a product strategy, because its business is focused on providing a wide range of products and services, rather than a single product or product line. As a technology company, Google offers a diverse range of products and services, including search, advertising, cloud computing, hardware, software, and many others.
Because of this diversity, it is not feasible for Google to develop and implement a separate product strategy for each of its products and services. Instead, its company strategy is likely to focus on broader goals and objectives that apply across the entire business, such as growth, innovation, and customer satisfaction. This allows Google to manage its business in a more comprehensive and cohesive manner, and to ensure that its products and services are aligned with its overall goals and direction.
The rant was interesting in it's time but I'm not sure how relevant it is in 2022. 11 years ago Google had a nascent cloud computing business and Amazon might have been at a bit of a peak in terms of executing on it's dual track cloud/retail businesses.
In the time since, consumer trust in Amazon has eroded massively, as China-based sellers turned Amazon into a more expensive Ali Express. Amazon has had some very public missteps in investments such as losing 10 billion a year on Alexa. Amazon's market dominance of the cloud has been eroded by Microsoft & Google. And even now, 10 years later, Amazon still can't seem to figure out how to make its retail business profitable without resorting to littering their web site with ads.
Amazon recently set a record for losing over one trillion dollars in share holder value, while it's competitors like Microsoft and Google have weathered the economic storm better. Microsoft and Google have very promising trajectories for their cloud businesses while at the same time having a solid profitable cash cow in their core businesses. Alternatively, one gets the sense that if Amazon didn't stumble upon AWS as a business all those years ago, they might legitimately be in trouble as a business.
> Amazon's market dominance of the cloud has been eroded by Microsoft & Google
By Microsoft. GCP appears to be under intense pressure to become profitable and they’re lagging on features. I hear about migrations from GCP to AWS or Azure more than the reverse, and that was before the price increases.
There are rumors about what's going on internally regarding GCP, but by all accounts Google is still investing heavily in GCP and their current growth trajectory means profitability is almost inevitable.
Azure & AWS have relatively stagnant growth and there are more and more signs that the growth of GCP is coming at potential market share from AWS/Azure.
I'm not saying it's a done deal but I think Yegge's rant holds up well. 10 years later, GCP has fallen behind Microsoft and is still a third of AWS' share. What would concern me more as a Google shareholder, however, is the relative feature set and progress — I use both professionally and GCP has been a real disappointment for how periodically running into basic functionality which is a checkbox on AWS but on GCP is a 2 year old issue in their tracker with no progress and a suggestion that you build/buy something and run it yourself. It's not like you can't use it successfully but it's hard to shake the feeling that one of these companies is a lot more committed than the other (which started having to cajole them to show up and sell their product).
More succinctly, I’ve recently seen a number of AWS announcements for things which make my life easier and/or save money, some of which I am already using. The last news I got from GCP was that cloud storage prices are going up.
The way I am reading that link, AWS + Azure both increased revenue from last year (33% to 34% and 20% to 21%)? The article says GCP currently increased from 10%->11%.
Which is not say anyone should stay content, but it looks like GCP still has a long road ahead.
Slightly beside the point, but for people who were already using Ali Express and wanting a faster, slightly smoother experience, this actually increased Amazon's value proposition many folds.
On the credibility part of AWS because of Amazon's overall image, I think the split comes at wether the focus is on renown brands and traditional items (image goes way down), or on items that simply don't exist or are incredibly cumbersome to find through traditional retailers (Amazon appears on the map as a viable Ali alternative).
I'm still playing the game of finding regular electric shavers that charge through usb-c only (no adapter, no charging stand). Traditional makers don't seem to give a damn about that segment and Ali / Amazon is the only game in town.
> Amazon recently set a record for losing over one trillion dollars in share holder value
I think is this case Amazon was massively overvalued, the market had unrealistic expectations
In the similar manner oil companies are overvalued - they seem to assume that every last drop of oil in discovered reserves can be burned, when in reality wars and social unrest will break out long before that.
Let employees to start new projects and shut it down later if it does not seem to be profitable. This strategy just make sense. But I just want google to make the works open sourced when abandoned. I hope google opensource Duplex for web...
By and large Google can't open source anything internal even if they wanted to, because it's so closely tied into Google's own ecosystem (Borg etc) that it's useless outside it.
>Let employees to start new projects and shut it down later if it does not seem to be profitable.
That's not the strategy though.
The strategy is "I'll get promotion for launching, not maintaining" for employees, and "we're still making a shit ton of money, so why bother" from the top brass.
Strategy leads to winning the big picture. And Google is not doing that. They are the King's of advertising, but move outside that stream and it becomes "when" not "if" it'll get the chop.
This leads to a lack of trust, so things like GCP suffer. Regardless of the technical merits, or price, I can never use GCP because I don't trust Google. This is largely due to their willingness to kill things, coupled with a lack of phone support.
Incidentally lots of the things shut down were successful, and profitable, by most company standards. Just not by google-scale standards.
I've never hear of anyone who loves Microsoft Excel. It isn't that it is a bad product (it is of course very good at what it does). It's just that spreadsheets are no fun.
Tried many different search engines and none was as good as Googles until I tried Brave search, something I really didn’t expect. We will see what happens with Googles search business in the coming decade.
The example the author gave seems flawed. Why not geofence that optimization to Norway? Or to Norwegian google? It’s a micro optimization but the search engine is full of those and google has 100k employees
guys we need to stop supporting these obvious bots. nothing they say matter. they won't even let people read this piece without some kind of subscription that goes towards their online social status.
For what it's worth, it seems authors can disable that popup on their own substack. I did, and it doesn't seem to have reduced subscription sign-ups, but even if it does, not annoying people is always a good principle.
It defaults to 'on' though, and maybe not everyone has explored their settings. It took a while for me to find it.
After trying Android on Samsung for the first time in a few years ... I feel it's a bit of a disaster. I could almost not find the 'settings' with the help of 5 other people. I was at a loss to know what I could delete and not. After 20 minutes of trying to delete an 'Amazon' app I realize that Samsung probably forced it to be pre-installed and not removable. Along with 10 other garbage Samsung and carrier icons. Paradoxically most of that does not even benefit Samsung.
This individual also had almost 10 terrible spyware apps installed without remembering that they had installed them.
Such poor design. Like an ugly, dirty esplanade off the main strip in some gambling resort. I'm awaiting the gambling and girls preloads.
I loathe so many of Apple's policies but I wonder if it's better to have a monopolistic overlord that is 'mostly benevolent' over the Android bad design and chaos.
The sad thing is a Trillion dollar company should have a better product.
I suggest Microsoft in many ways is not that that much better as can be seen with the abysmal degradation of Skype, and MS Teams which is a poor copy of Slack. This could be a function of the nature of those divisions, or of the company losing it's touch.
> After trying Android on Samsung for the first time in a few years ... I feel it's a bit of a disaster. I could almost not find the 'settings' with the help of 5 other people.
Is this some kind of AI generated comment? It's in the app list under settings, also in notification shade next to search button, and first result in search.
Five medical nurses in a 'good' Hospital could not find the setting.
So tell me where the 'joke' is?
(and frankly it doesn't matter what 'phone' background they have.)
The 'joke' are those (especially designers) that completely lack the self awareness to grasp their specific and narrow perspectives are not normative and therefore do not necessarily represent a 'Good UI'.
In particular, UI elements that change over time, with every prouct iteration, putting new feature in new locations. Etc..
After having to deal with that and struggling to remove several spyware apps - and unable to remove bloatware apps - it's clear that Android is a second class OS. And FYI it does not matter that Samsung gets it for free - it's Android.
It's the tip of the iceberg, and aside from side-loading and price, I don't see how this thing is in any way competitive with iPhone.
People defending 'Shop At Amazon' planted on their home screen have lost perspective.
If you had never used an Iphone you'd probably struggle to find settings as well. Your complaint seems to be "The user interface is slightly different from the Iphone."
IMO that's a good thing, 99.9% of people who try to delete the Settings app probably shouldn't. For the average user it would be basically a bricked phone at that point.
And Samsung lets you easily search for settings, just like IOS does. It's extremely easy to find them unless it's an odd feature with an ambiguous name.
> I loathe so many of Apple's policies but I wonder if it's better to have a monopolistic overlord that is 'mostly benevolent' over the Android bad design and chaos.
The problem is benevolent dictators don't stay benevolent. Or, to put it differently, things are great while you're on the same side as the dictator, but, once you get on their wrong side, things get bad suddenly, and it's hard to find allies since your erstwhile community is probably still on the right side of the dictator.
(I am not a software developer, much less an iOS developer, and have no experience with being on the wrong side of Apple—but I do have experience, as a long-time OS X user who now finds macOS becoming "iOS, but less important!", with what happens when the dictator controls your working environment but doesn't care about it any more.)
You’re right. Apple is taking the Mac less seriously. It’s not like they just designed an entire line of processors just for the Mac and created laptops that are the fastest in the industry with three times the battery life.
> You’re right. Apple is taking the Mac less seriously. It’s not like they just designed an entire line of processors just for the Mac and created laptops that are the fastest in the industry with three times the battery life.
The Mac hardware is still taken seriously. The software is not, in the sense that the distinction between iOS and macOS (which I value) is being—not just negligently but intentionally—erased.
(To be sure, there are people for whom that's a good thing, and maybe even objectively, on average, it's a good thing. But I think it's hard to argue that Apple as a company cares about what OS X used to be.)
> After trying Android on Samsung for the first time in a few years ... I feel it's a bit of a disaster. I could almost not find the 'settings' with the help of 5 other people.
Is this a joke? It's in the app drawer, it's called "Settings".
Not even hidden in a folder, it's right there. And if you're that blind, the app drawer also has a search box... Settings also has an icon prominently placed in the notification tray. And the settings app itself also has search so you can search for whatever setting you need. You can also search "Settings" in the Google app/search bar widget, it brings up the app as well.
Like it or not, Samsung is the premium phone in the Android ecosystem. That is cheapens the experience with bloatware is a real blight on the ecosystem. Your options are either Pixel or Android One. Last I looked, few models are still utilizing the Android One program.
Recent, begrudging convert to iPhone, because Google refused to fix 911 calling.
There is nothing “open” about Android in a practical sense. Most of what makes “Android” Android outside of China is closed sourced Google Play Services and related apps and Google has requirements as far as when a manufacturer is allowed to use the closed proprietary parts.
> To support this company strategy, Google makes it really, really easy for anyone to start up a new project. The culture of “20% time” gives people permission to work on ideas outside of their primary role. There’s only the lightest of reviews needed to launch a “1% experiment.”
No, it is not super easy to start a new project. I'm in Google Cloud and we're pretty focused on executing team priorities. I can suggest new work-streams and write arguments in their favor, sure. But I can't just randomly add a feature or go away to work on a new product for a while. Every public thing in Cloud, at least, is carefully reviewed with a large checklist covering privacy, security, accessibility, API style, integration with existing systems like metrics, legal, etc. The simplest cloud service can take an extra six months or a year to get past these blocks. These requirements also apply to features. What we get for this is quality and consistency between products/features.
And 20% time isn't really a thing any more. You can get permission from your manager to work on something else, perhaps. But more likely this sort of thing is more like 120%. It's not at all as lackadaisical as it's presented to be. The most I see around my team is people volunteering for large-scale clean-ups. But these are more slack-time projects like 5% projects. It's not totally gone however. I think the person who made Google Maps zoom out to a globe did it as 20% project, and I believe adding CitiBike support was similar as well.
We are also super careful about experiments. Every Google Search experiment has to go through a senior management review. In my team, every flag rollout is discussed and approved before rollout.
Insofar as these tidbits are presented as evidence for loose leadership, they don't make sense. The ship is fairly tight, at least in Cloud.