Alphabet earnings show Google Cloud on $10B run rate
techcrunch.com
techcrunch.com
Google utterly fails to understand the need for its customers to trust that google will support and service them and not destroy their business after building on a google platform.
Apparently they are completely oblivious to these things being important though.
Right here on HN are the influencers who tell companies to use this or that techonology and there’s a loud howling from everyone that googles support is beyond bad and that you’d be mad to risk building on any google service cause they’ll shut it down.
Google has internet in its DNA it never did anything else but internet.
But it failed to see the opportunity in the first place.
Secondly I attribute its lack of customer service to its recruiting and hiring practice.... it only ever hired the most geeky of geeky academics.
And geeky academics are the opposite of human relationship oriented.
Thus everyone at google designed the company to avoid human contact.
And that surely the complete opposite of customer service.
Google has the inverse of good customer service DNA... it has “hide from and avoid customer contact” DNA.
And that’s what has sunk the google cloud. $10B might sound like a lot, but to me it sounds like a gigantic fail.
It's impressive to me that Amazon was able to shift focus to AWS without much existing presence in enterprise tech stacks.
Still though, enterprise is a tough domain in my experience, AWS did a great job
That's the new trend though for almost all companies. Customer telephone support is costly, difficult to manage, difficult to staff, and problematic customers consume an overwhelming amount of resources - orders of magnitude more than their value as a customer.
...on the enterprise side, I've had great experiences with their GSUITE support teams.
The online support forums are garbage though - I'll give you that.
A healthy company mines the hell out of it for feature discovery and product improvement.
These people are literally continually talking to your customers about what their actual pain points are!
I mean, what kind of support does a small business with two EC2 Instances or three Office 365 licenses really get? I can’t imagine you get to call someone and get in-depth technical help for free.
In other words, I think you’re repeating really common criticisms of Google in general, but I can’t tell if you’re speaking objectively or from a matter of pure opinion.
That small business can also look at https://aws.amazon.com/simpledb/ and know that a 15 year old boring and low utilization service still bums along.
One could argue also that AWS has too many often esoteric services, and if they focused on making some of them more feature-full the service would improve (disclosure we use both and I think both are perfectly good cloud services).
No, I don't think there really is.
With AWS, I've got full confidence that if I engineer something on a service like SimpleDB, it will keep chugging along on the AWS side long after my side stops working. I don't have that same confidence in Google.
AWS frequently will communicate detailed, specific technical workarounds and instructions via email on the free support plan. I’ve reached out to GCP support and the best I’ve ever received was a link to a generic support document.
AWS does give 1-1 support to smaller clients, and you can also pay extra to get actual phone help. The OP may or may not have had direct experience with this but his criticisms ring true with mine.
Give me AWS support any day
On top of that AWS documentation is often both needlessly opaque, elliptical, incomplete, and outdated or otherwise incorrect.
On the other hand, if the issue isn't too obscure, AWS’s huge marketshare means that you can usually find decent answers on SO.
You essentially have to be a systems engineer/sysadmin for every one of your customers with only as much context as your customer is willing to share.
I like my job, but please have mercy on support.
The reason Google support has a hard time of it is Google doesn't offer that model to all its customers, just the ones that can pay a lot of money. But support still has the job of helping everyone else (with all the challenges you've described that such entails).
Their zeitgeist appears to be "Other companies needed to provide that because their technology was wrong / incomplete. We'll just build things right instead."
Which is batshit insane, in the same way that expecting a veterinary pharmacist could prescribe for humans... with better technology.
Contrast that with AWS, where I had 4 engineers (back when our bill was 12k/year) helping us with recommending certain arquitectures and actively being useful anytime we reach out.
Our AWS expenses tripled because of growth and us choosing them for new components while our GCP expenses have remained flat.
Amounts like 30k/year may not sound big to many folks, but when you go to a truly good SaaS organization and spend that kind of cash, you get excellent support (online or offline). Google has absolutely amazing services that for the most part do not require human support, but for the times that you actually do, it is simply non-existant.
It may be because we're an attractive customer with a very large amount of AWS spend, but the folks at GCP have made our adoption pretty amazing. We've had multiple on-boarding and troubleshooting calls with the actual product managers of GCP services, in addition to onsite training/consultation with experts (all paid for by Google).
I've had to use support a couple of times and found it pretty on par with support I've gotten out of AWS.
Like any provider of services, I've found that relationships are key. Cultivating a good relationships with your reps/contacts goes a long way in their willingness to go the extra mile for you.
But, interactions with our reps have grown fewer (we're told their customer success/technical account management teams are understaffed), and it may be a sign that we're getting out of the honeymoon phase. We'll see.
As have we. But that should ring a few alarm bells in regards to scalability and whether that can continue.
Google should just cut IBM's support org (the older, experienced folks) off the carcass, retrain them, and use them as a cadre for building their own org.
The point is that that doesn't matter. Even if it's not always logical, the reputation of your company as a whole sometimes matters, even in the case of corporate accounts ("nobody got fired for buying IBM" after all). Moreso in the case of startups where you have one tech guy who might have gotten burned before.
Amazon has a stellar reputation since nearly everyone has gotten a refund from Amazon for some cheap DOA $2 cable but meanwhile Google stonewalled them on some random glitch in Gmail. Even if it's not logical, those things stick around in people's heads.
Very communicative, real people responding...couldn't have asked for more. Also I personally feel the UI on GCP is far more human-centered/focused than either AWS or Azure.
To me, GCP is the most "human" cloud service offering so far. I get the criticism when it's applied to GMail or Google Maps but I really don't understand that as applied to GCP.
I completely disregard any documentation or information older than 2019, and I'm a bit suspicious if it is from the first half of the year, that is how often they change things for no apparent reason.
Their support will likely try to convince you to rewrite your app with their latest stack than solve obvious issues with Azure.
The only positive thing about the whole ecosystem is that support and evangelists/presales will likely admit bugs and share workarounds they've used. For example deploying AKS from the portal has a 50% chance of failing because it will not have created the service principle before something else requires it - so they will recommend creating the service principal in advance and selecting it in the config.
Like Valve they have created products that took over the whole space.
They have a tendency to completely drop things when they don't pick up traction early enough. (artifact, everything Google)
There used to be a Halo around both companies in the 2000s, that was slowly eroded in the 2010s.
They suck at communicating.
They create good products that never see the light of day again (Portal, l4D,Half Life) because their main product (steam, ads) makes so much more money.
________
GCP has the same problem as Artifact, where Google bull headedly tries to enter a different type of product/business model (b2b, instead of b2c) without adopting any of the cultural must-haves of a b2c company. (customer before product, service, reliability over speed)
Tensorflowv2 is failing due to a similar kind of stubbornness.
_______________
The scary thing is, if Google manages to pull off the culture change (unlikely), I can see them sweeping both AWS and Azure in the cloud space.
Contrast this with the other cloud providers who seem to be using their own clouds for their primary services creating a feedback loop which ensures they don't have the big outages that GCP has had over the past year.
This is something only Valve could do, because Steam makes so much more money. No other company would leave such a lucrative franchise for so long.
Meanwhile, perpetual testing initiative was a great solution for keeping the game active without requiring active investment from Valve's game developers.
You certainly could run on it on a dedicated combination of any mix of wind/solar/hydro power and stored (e.g., battery or regenerative fuel cell) power you wanted to.
In practice, you'll probably want to connect it to the grid, but directed purchases on the grid, while they don't actually select which source really powers the DC, have all the practical effect of doing so.
Only somewhat related, but I'm quite certain that on premise solutions would win over most cloud services if the energy cost of data transmission was factored in.
But if you do that, I'm not sure what would even differentiate them. They're all using essentially the same hardware. They all have the same cooling requirements. I wouldn't expect an average to be very interesting.
Distance from carbon-neutrality per dollar of revenue sounds like a good start.
Keep it simple, just publish how much power they're using and what the sources are.
However: one could say that these companies are now big enough that it is reasonable to start demanding that these companies source the energy in a climate-friendly way. I think it's not at all unreasonable to demand that when Google or Amazon builds massive new data centers to be responsible about how those data centers are powered.
Google reports an average PUE[1] of 1.11 over the past year, 1.09 over the past quarter, and the latest/best data centers are at 1.06 [2]. In simple terms this means that the total power overhead for cooling etc is just 6% of that to power the machines in them.
Google (and others) have gone to great lengths using AI, and radical new ideas for cooling.
Disclaimer: I work at Google, but not in Datacenters. All info public domain
[1] Definition of PUE: https://en.wikipedia.org/wiki/Power_usage_effectiveness
[2] Source: https://www.google.com/about/datacenters/efficiency/
Google regularly brags about their facilities and you can find info with 30 seconds of research. AWS is a little more cagey, they will brief customers. Microsoft is more like Google.
https://www.blog.google/outreach-initiatives/sustainability/...
...Microsoft is setting the goal of being carbon negative (since the inception of the company in 1975): https://blogs.microsoft.com/blog/2020/01/16/microsoft-will-b...
Edit: according to other comments Microsoft's numbers are no longer inflated.
The rest of it is from Windows Server and other server products like SQL Server, SharePoint ... and Enterprise services. As an example SQL server running in AWS also counts in that number.
<quote>
"Revenue in Intelligent Cloud was $11.9 billion and increased 27% (up 28% in constant currency), with the following business highlights:
Server products and cloud services revenue increased 30% (up 32% in constant currency) driven by Azure revenue growth of 62% (up 64% in constant currency)
Enterprise Services revenue increased 6% (up 7% in constant currency) "
</end quote>
https://www.neowin.net/news/microsoft-q1-2020-earnings-reven...
Office is not counted in cloud. It’s counted in “Productivity and Business Processes”.
That...doesn't even make sense. He premise doesn't even have a rational connection to the conclusion.
I understand that this is probably legal under the current system but there have been plenty of calls to change that system lately as it has been heavily pushed to its current shape through decades of lobbying & special interest. It's not normal for corporations to dwarf most of the nations in the UN in terms of GDP. That's not a free market anymore but a form of corporate dictatorship.
Wishful thinking, these people will earn money regardless.
I don't know about that. As someone who has worked for some of the big companies, there's a ton of efficiency gains in being able to use a common set of internal tools and software that these companies have and while the public clouds have externalized a lot of those internal tools, they haven't externalized all of them yet nor would you benefit from the common set of software reuse.
It has been, actually, normal for there to be corporations with revenue dwarfing the GDP of most nations on Earth (the UN is a recent distraction) for nearly as long as joint stock corporations have been a thing.
Largely because GDP isn't exactly equally distributed among countries; the median GDP of a country on Earth right now is only in the neighborhood of $15 billion, which is in the neighborhood of the GDP of a quite small city in the developed world.
> That's not a free market anymore
“Free market” is a non-existent abstract ideal, not a thing that actually exists or can exist.
How does a small cloud provider stay in business when up against these multi billion dollar companies? At any point in time, these companies can just lower the cost of their products, and suffer through a few bad quarters while the small businesses slowly bleed out and die.
> while the small businesses slowly bleed out and die.
This is what killed small grocery businesses. The bigger corporations lose money on lots of popular items and make margins on other products. Walmart is especially bad, they can take lots of losses on grocery because they're making margins in other areas of the store.
The ultimate question is what is allowing these bigger businesses to operate more efficiently than a small business? Offshoring labor forces, offshoring manufacturing, low to no import tariffs, ability to utilize immigrant labor (small companies have no shot in the H1B process), corporate HQ's in lower tax jurisdictions, legalized accounting fraud (CapEx vs OpEx), skirting labor laws by hiring 'contractors,' getting outright handouts from municipalities (see Amazon HQ2 or any NFL stadium), onerous regulations, etc, etc.
Democracy in action. This is what everyone keeps voting for. Kill small business, empower multinational corporations.
I would think technology allowing businesses to scale with extremely low, near zero, marginal costs would be the big reason. That's why the per employee net income figures are so high for the leading tech companies, they can do so much with so little. I predict lots of vertical monopolies in the future.
What's amazing is that, literally nobody goes to the polls hoping to vote for such outcomes, yet despite that fact, it is indeed what we end up getting. Democracy is simply unable to fix this.
And why would they even need to do something like that? AWS is wildly profitable anyway with enormous amounts of market share. When would it ever make sense for them to give up billions and billions and billions of dollars in profits to wipe out tiny cloud providers while also screwing up the economics of the space (changing people's expectations of how cloud should be priced meaning they won't accept prices as high as they are currently) and hoping and praying that GCP and Azure play along when it's time to hike prices back up?
This behavior is illegal yet Amazon is still here. While there’s no obvious case for their cloud division, it’s not unprecedented within the company to do so.
https://www.nationalreview.com/corner/misplaced-trust-antitr...
It's not a guaranteed formula---AWS, Google, even Azure have big moats around their offerings and significant table-stakes features implemented. But "focusing entirely on the needs of the customer" is the thing their scale makes challenging; they have customers with multiple disjoint or competing needs.
What could be contributing to those large numbers?
[1] - Memorystore only allows vertical scaling, no support for GCP managed Redis clusters, no hybrid functionality (only instances in the same VPC network can access Memorystore, whereas ElastiCache offers all of this.
https://cloud.google.com/memorystore/docs/redis/networking https://cloud.google.com/memorystore/docs/redis/scaling-inst...
https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/... https://docs.aws.amazon.com/AmazonElastiCache/latest/red-ug/...
[2] - https://cloud.google.com/memorystore/ vs. https://aws.amazon.com/elasticache/pricing/
Disclaimer: I work at Google Cloud.
Are they really?
https://www.parkmycloud.com/blog/aws-vs-azure-vs-google-clou...
Clearly Cloud market in general is growing and all 3 companies want a piece of the pie.
If they can all make a sizable business from the growth it doesn't matter who is growing the fastest or who is the largest.
AWS and Azure are both much better about launching new pricing structures, and generally when they make something more expensive its because they're adding new features (and corresponding price tiers).
I disabled uMatrix in a Private tab but my pihole blocked advertising.com too.
I nearly gave up thinking "I don't need to see TechCrunch, their loss" but then I remembered the wayback machine:
https://web.archive.org/web/20200206070011/https://techcrunc...
The article loads but after a few seconds it auto forwards to a page that doesn't exist on the TechCrunch site. I get about 3 seconds to read it.
TL;DR: Google cloud revenue includes Saas offerings (docs, gmail etc), in addition to the infrastructure part. They may make roughly 2.5bln per quarter now, but it's still small compared to Microsoft (12.5bln per quarter, includes Azure + Office/outlook) or Amazon (10bln per quarter, AWS only). What is impressive is the growth of the business unit - more than 50% in the last year (they do have to keep it up at the same rate for 4 more years to catch up to competitors though)
This trick apparently still works: after the page loads and before it disappears, hit ESC. Instant-freeze and you can read the article! :)
It improves the user experience by a lot on Techcrunch, medium and a bunch of other news sites.
Cookie banner are not directly related to the GDPR, but to the Eprivacy directive, which, not being a Regulation, is subject to variation in national law transcription. So the answer may vary depending on your country!
Still in an effort to have a uniform position on the question, DPAs( Data protection agency) are working at the EDPB (the European Data Protection Board, a sort of council of DPAs) to have a common position.
One of the currently shared position is that a cookie wall impedes the "free" part of the consent, since you have to support a negative consequence.
In a communication regarding the (future?) Eprivacy regulation, the EDPB clearly stated that:
"In order for consent to be freely given as required by the GDPR, access to services and functionalities must not be made conditional on theconsentof a userto the processing of personal data or the processing of information related to or processed by the terminal equipment of end-users, meaning that cookie walls should be explicitly prohibited." [1]
If I remember correctly the Austria DPA is not of favor of a cookie wall ban though..
From memory, not a legal advice!
EDIT: try to make the explanation clearer!
[1]: https://edpb.europa.eu/sites/edpb/files/files/file1/edpb_sta...
> ‘Personal data’ means any information relating to an identified or identifiable natural person (‘data subject’); an identifiable natural person is one who can be identified, directly or indirectly, in particular by reference to an identifier such as a name, an identification number, location data, an online identifier or to one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of that natural person.
If the tracking id cannot be correlated to a name, identification number, precise location data (not country level), then it's not PII as far as the law is concerned. The criteria is clear: "relating to an identified or identifiable natural person". There is no way that simply a session ID stored in a cookie can be traced to an identity IRL.
I fell that I know what I'm talking about as I designed and implemented an customer authentication system for a medium-sized company that is based in EU, needs to respect GDPR, and I worked closely with their lawyers and operations to make sure we are fully GDPR compliant, and we passed the relevant audits.
Which is still not PII. More importantly, 3rd party advertising cookies CANNOT be linked to a user account if you don't have code that stores them in your environment. CAN has a very limited meaning, whereas it requires all the preconditions to be true (i.e. I'm storing both 3rd party cookie ID, AND session ID, AND the tables have a correlation), it does not mean "COULD if more code was written".
> and from there someone's name, address
Only if you ask for AND store those. If you're asking for example for a real name and address for an e-commerce transaction, and you're passing them to the card processor, and not saving them anywhere, not even in logs, then you're not storing PII, and you CANNOT link tracking cookies and session ID to data.
Don't get me wrong, I'm still using Firefox containers, and uBlock Origin, and pi-hole, so I totally don't like to get tracked, even if anonymous. But the tendency on HN to label anything that could be used to track a user as PII is actually damaging, because it creates false expectations about how the law actually works and how much somebody is protected.
It doesn't really matter though because literally nobody it enforcing this part of the GDPR.
Between us, I’m not sure why they are so addicted to various tracking that tells them that absolutely no one clicked on 90% of their content, but they are, and they lack the technical ability to do it themselves without relying on frameworks that steal privacy information.
But if people really did overwhelmingly say no, I just see no way for most of the internet to exist. You get overwhelmingly less per click/impression for 'dumb ads,' and news sites have already had to resort to click bait today. It'd pretty much guarantee anything not owned by one of the top 10 would be paywalled in some way.
https://www.lexology.com/library/detail.aspx?g=1cad75f1-5fe1...
Edit: On second thought, I'm not sure if the cases are sufficiently similar as Der Standard offers a subscription free of tracking and advertising as an alternative for those who reject cookies. I'm not sure if buying a TechCrunch subscription means they will stop tracking you.
For now GDPR seems like one more toothless EU regulation. Maybe they manage to catch some big US company and make them pay a few millions just to make a big show and justify the existence of GDPR. Maybe they will catch some poor guy maintaining forum for some hobby group and does not provide "right to be forgotten" functionality (that's why effectively all independent forums are going away in favor of Facebook - good job, EU). But I doubt GDPR will manage to give people more privacy.
I hope that people themselves will figure out what's going on and start fighting back (by using browser plugins, stubbornly reporting misbehaving sites, trying to engage authorities to enforce GDPR, etc.). If yes, maybe GDPR will turn out to be something valuable, for now, it is not.
https://chrome.google.com/webstore/detail/cloud-to-butt-plus...
Why do they do it?
Because AWS is at a $40B run rate, and both Google and Microsoft have to show analysts and customers that they are catching up, and that they are big enough, etc.
Real "cloud infrastructure" revenues for Google are probably less than half that.
The worst part is that famed and well-paid analysts perpetrate the lie, either by collusion, or by ignorance.
Edit: Apparently Microsoft is already doing this.
>Turning to Microsoft, it reported a combined cloud revenue, which includes SaaS (Office 365, Dynamics, etc.) and cloud computing (Azure), of $12.5 billion for the quarter.
Could you provide a source? I can certainly believe TC is wrong, but contradictory information isn't great.
"Revenue in Productivity and Business Processes was $11.8 billion and increased 17% (up 19% in constant currency), with the following business highlights:
· Office Commercial products and cloud services revenue increased 16% (up 18% in constant currency) driven by Office 365 Commercial revenue growth of 27% (up 30% in constant currency)
· Office Consumer products and cloud services revenue increased 19% (up 20% in constant currency) with continued growth in Office 365 Consumer subscribers to 37.2 million
· LinkedIn revenue increased 24% (up 26% in constant currency)
· Dynamics products and cloud services revenue increased 12% (up 15% in constant currency) driven by Dynamics 365 revenue growth of 42% (up 45% in constant currency)
Revenue in Intelligent Cloud was $11.9 billion and increased 27% (up 28% in constant currency), with the following business highlights:
· Server products and cloud services revenue increased 30% (up 32% in constant currency) driven by Azure revenue growth of 62% (up 64% in constant currency)"
So Office 365 falls under category "Productivity and Business Processes", while Azure as well as 'server products' and 'enterprise services' fall under "Intelligent Cloud".
Their third category is "Personal Computing" which was $13.2 billion and includes Windows, Surface, Search / advertising (I guess Bing), Xbox.
Edit: Does this include SPLA licensing? Trying to figure out what all is in there besides Azure.
Azure the PaaS and associated cloud offerings are much less.
The root of the problem is that MS decided to expand what "Azure" is for marketing and financial reporting purposes beyond its PaaS origins.
Azure is under “Intelligent Cloud”, which earned $11.9b this last quarter.
0: https://www.microsoft.com/en-us/investor/earnings/FY-2020-Q2...
AWS has $40B annual run rate (https://www.zdnet.com/article/aws-brings-in-nearly-10b-in-sa...)
Something is not quite adding up.
Is there any public information about it available?
https://www.neowin.net/news/microsoft-q1-2020-earnings-reven...
I personally prefer things like digital ocean or even heroku becaus they are so much easier to use and manage, but it’s really hard to build a business case on non-azure when you have 365, and there is no alternative to 365 if you’re in a GDPR sector.
We have plenty of procurement projects that are hosted in AWS by private suppliers, so it’s not like we’re somehow against other cloud vendors, it just doesn’t make sense to not chose Azure when you have 365, because that means you’re already partly there and already have the Microsoft certified and trained it staff.