Google Execs Debated Getting Out of Cloud Computing Last Year, But Instead Set a Goal of Being a Top-Two Player By 2023 (cnbc.com) 74
In early 2018, top executives at Alphabet debated whether the company should leave the public cloud business, but eventually set a goal of becoming a top-two player by 2023, according to a report from The Information on Tuesday. From a report: If the company fails to achieve this goal, some staffers reportedly believe that Alphabet could withdraw from the market completely. While Alphabet subsidiary Google is dominant in web search and advertising, the company is still a small player in cloud computing, which involves renting out computing and storage resources to other companies, schools and governments. In 2018 the company lagged Amazon, Microsoft and Alibaba in that market, according to industry research firm Gartner.
it would be sad if they left (Score:2)
Re: (Score:1)
Their offering is a lot more user friendly than AWS, and a lot more feature-packed than Azure.
Rephrased: Their offering is a lot more dumbed-down than AWS, and a lot less feature-packed than Azure.
There, fixed that for you.
Re: it would be sad if they left (Score:2, Insightful)
You said "Dumbed down" like that's a bad thing.
You can offer as many features as you like, but what's the point of they are too complicated for people to realistically use.
Personally, I would rather have fewer features that I actually used and are efficiently implemented then hundreds of features I don't use which make everything overly complicated and inefficient.
But you be you.
Re: (Score:2)
Re: (Score:1)
Well, to acknowledge a bit of reality, most of Azure is Office 365. And the rest is companies who are moving Windows applications 'to the cloud' via Citrix - which only Microsoft can offer without tacking on Windows licensing fees. And besides, Citrix - yuck! But hey...
Re: (Score:2)
That's no where near reality.
https://azure.microsoft.com/en... [microsoft.com]
Windows Virtual Desktop is a tiny tiny offering. Most of the Azure stack is now PaaS components you use for building applications (much as AWS). Even IaaS is only a tiny part of their offering.
What all there do you think is tied to 365?
Re: (Score:2)
Companies that use Outlook are being actively encouraged to move their exchanges to Azure. That probably accounts for hundreds or more of Azure 'installations'.
The rest I guess I don't have evidence for - except that the company I work for has chosen Azure to host their old fat-client applications 'in the cloud'. And what that amounts to is Citrix. Now, granted, that's a stopgap measure. The stated intention is to rewrite those apps to be cloud-native, but those efforts are way behind schedule. Thousa
Re: it would be sad if they left (Score:5, Interesting)
You said "Dumbed down" like that's a bad thing.
You can offer as many features as you like, but what's the point of they are too complicated for people to realistically use.
Personally, I would rather have fewer features that I actually used and are efficiently implemented then hundreds of features I don't use which make everything overly complicated and inefficient.
But you be you.
I've been playing around with Google Cloud's offerings and I've come away very impressed. The interface is remarkably easy to use, but the level of documentation for non UI functionality is astronomical.
With AWS, everything has a UI and can be done with through the web interface or you can dabble behind the scenes. It's up to you.
Honestly, I'm glad Google didn't jump ship. I'm anti-Amazon for pretty much everything from the store to their server architecture.
The more cloud players the better.
Re: (Score:3)
I've been playing around with Google Cloud's offerings and I've come away very impressed. The interface is remarkably easy to use, but the level of documentation for non UI functionality is astronomical.
With AWS, everything has a UI and can be done with through the web interface or you can dabble behind the scenes. It's up to you.
Honestly, I'm glad Google didn't jump ship. I'm anti-Amazon for pretty much everything from the store to their server architecture.
The more cloud players the better.
You only think that because you have only been playing around. As soon as you need some measure of reliable behavior, GCE isn't for you. I'm sure it works for Google's offering but Google's use cases aren't anything like yours or mine. Those differences really start to show when you start moving prod to GCE. Error rates are far higher because Google's default behaviour is to issue all requests twice and cancel the one that either breaks or finishes second. That's fine for Google, but a nightmare for yo
Re: (Score:3, Insightful)
Their offering is a lot more user friendly than AWS, and a lot more feature-packed than Azure.
Perhaps, but the market doesn't see it that way.
A rule of thumb in businesses is that the leading company makes 60% of the profit, number two makes 30%, and everyone else fights over the scraps.
So if you aren't #1 or #2, and you have no plausible path to get there, you should get out.
The problem is that by blabbing about their strategy publicly, they look unreliable, and even more of their customers will abandon them.
I'd think they'd have it for internal anyway (Score:2)
One would think Alphabet / Google has a big enough infrastructure and enough engineers that they would use a significant amount of cloud themselves. At this point, maintaining a private cloud for their own use is probably a significant cost savings versus switching to Amazon. Amazon's cloud has some cool stuff, and it's costly for the hardware you get.
Assuming they aren't going to completely shut their cloud down, it seems reasonable to allow their large customers to keep buying from them, even if they are
Re: (Score:2)
One would think Alphabet / Google has a big enough infrastructure and enough engineers that they would use a significant amount of cloud themselves. At this point, maintaining a private cloud for their own use is probably a significant cost savings versus switching to Amazon. Amazon's cloud has some cool stuff, and it's costly for the hardware you get.
Assuming they aren't going to completely shut their cloud down, it seems reasonable to allow their large customers to keep buying from them, even if they aren't the second largest cloud provider in the world.
Except I think their private cloud (for hosting public services like mail, search, etc) and their public cloud business do not overlap.
Re: (Score:3)
too bad they don't care about money (Score:2)
but only about what ranking they have in size.
frankly thats an insane way of doing business.
Don't use Google's cloud (Score:5, Insightful)
If the company fails to achieve this goal, some staffers reportedly believe that Alphabet could withdraw from the market completely.
In other words, people would be fools to use Google's cloud, because like so many other Google products, they might drop this one, too.
Re: (Score:2)
That's really become a liability for them... You don't want to become too dependent on any Google service... it might go away tomorrow ... I guess Android and ChromeOS are safe for now.... Who knows??
Re: (Score:2)
I wouldn't bet on either.
I suppose Android has the advantage of shorter usefulness cycles (phones often break).
But isn't Google already working on a new Mobile OS to replace Android? Also, the relationship between ChromeOS and Tablets is weird, and I could totally see them dropping it when whatever arrangements they currently have with schools are over.
Re: (Score:2)
Re: (Score:2)
The services that might go away from Google are called "BETA" like GMail was stuck in for a few years.
Re: (Score:2)
Hangouts, notebook, reader, google+, wave...
Re: (Score:2)
https://support.google.com/chr... [google.com]
Re: (Score:2)
And that is a bad thing? You do realize that ext4 only exists because the Lustre guys where wedded to ext3 and it would only go to 16TB.
Why 10 years ago would you want to use ext4 over XFS is beyond me. I am sure someone at this juncture will say something about XFS truncating files on a crash. Well I have news for you early versions of ext4 did exactly the same thing for exactly the same reasons till they deployed the fix that XFS had many years earlier.
So you use brand new code with lots of bugs that stil
Re: (Score:2)
I should have added that now the issues with using exFAT on Linux are going away, formatting external drives to ext4 is probably not the best of moves.
Re: (Score:2)
I currently i have my air-gapped storage running on a RPi4 and a 8 TB external formatted to Ext4. What format should i be using instead?
Re: (Score:2)
"Android and ChromeOS are safe for now"
If they stopped supporting them tomorrow, phones and laptops with that OS will still work. Now if they decided to shut down gmail tomorrow, many people and businesses would be in a world of hurt though.
Re: Don't use Google's cloud (Score:2)
That's true for all companies... ... But especially true for Google.
They basically said "you probably shouldn't invest money in us because we are probably not going to support you in the years."
Re: (Score:2)
That's true for all companies... ... But especially true for Google.
They basically said "you probably shouldn't invest money in us because we are probably not going to support you in the years."
I used to work in AWS. They had a firm rule that you can't take away anything anyone is using. You can't even change an API in a backwards-incompatible way. Any company can change its rules, of course, but the difference in mindset was stark.
And Microsoft has bet the farm on Azure, so I don't see it going away any time soon. MS has deprecated and then abandoned tech people were using, like the old VisualBasic, but they do it surprisingly rarely, and I don't think they've done it in Azure yet.
Maybe it's
Re: (Score:2)
First of all, MS does that with literally all of their software. The support it for a time, then they cut support. The software doesn't disappear from your hard drive, but the support ends.
Web services is far too young of an industry to make any long term predictions. Ultimately, I bet it ends up having features cut faster than most tech industries. The problem is support and reliability. If feature X is used by too few people and costs too much money to service, then it will get cut. Further, eventually,
Re: (Score:2)
MS became successful because 16 bit Windows 3.1 drivers (and any user mode software) would run natively on 32-bit Windows 95. They kept that shit working until the death of the non-NT Windows line, and even then most 16-bit usermode stuff worked until the 64-bit world arrived.
MS is two CEOs down the line now, of course, with a very different culture, but they do have a track record of trying pretty hard to keep your old shit working for many years. I wouldn't build a new business on Azure, but I can't say
Re: Don't use Google's cloud (Score:2)
Don't get me wrong, I don't think MS is going to kill Azure any time soon.
On the other hand, I have actual money bet in the office on MS functionally killing Windows. By Jan 1 2023, I bet that MS will have at least announced a plan to depreciate Windows by open sourcing major components, like Win32, and then rolling the Windows brand into a Linux distribution. This is primarily because historic support on Windows is unsustainable, and they want to migrate their business to services.
Re: (Score:2)
Re: (Score:2)
ROFL, can I have some of what you are smoking? Windows is a billion dollar business, no way in hell they are gonna go FOSS and toss all that revenue in a dumpster.
Well it was. Not anymore. Desktop/laptop sales are crashing (they are 20% of their 2007 high). That's why they are starting to think about an exit strategy. Most of Azure these days is Linux. Hell, when the MS saleman came to our office he spent about 15 minutes trying to get a Windows instance spun up in the cloud, then quit switched to Ubuntu to finish the demo. He said, its been months since anyone requested a Windows demo. They see the writing on the wall and have said they are getting out. Win3
Re: (Score:1)
Win32 will become a compatibly layer on a Linux kernel.
This made me imagine Microsoft contributing code to perfect WINE's Win32 layer.
Absurd? Stranger things have happened.
Re: (Score:2)
Re: (Score:2)
Too bad Hyper-V (both the old and the new Azure redesigned one) has no accelerated graphics support.
Re: (Score:3)
Microsoft are exhibiting a higher turnover of features and a more "We're turning this off, deal with it" attitude when it comes to their O365/Azure offerings than on-prem.
To an extent, that makes sense - they have to protect their systems that are running the software, and they don't want to deal with three thousand custom versions of each product for various feature sets in their production environments.
Still - they're doing things like killing support for standard POP3/IMAP to Exchange with about a 1 year
Re:Don't use Google's cloud (Score:4, Insightful)
Google's motto: Iterate Fast, Lose Interest in Things.
Re:Don't use Google's cloud (Score:4, Interesting)
A perfect example is Google's HQ in Mountain View.
While the grounds are maintained, all the lawn art and furniture, all outside decorations are decaying and rotting away. All the Android statues except the latest one are tilted, broken and missing chunks. And each new version of Android has a smaller statue that the previous one.
And of course, the official Google Visitor Center is permanently closed and the Google Gift Shop is open only 10-6 on work days - dealing with customers was never Google's strong point.
Re: (Score:2)
Oh, my sweet summer child. Surely you didn't think the unwashed masses were Google's customers, did you?
Re: (Score:3)
Exactly. The reason why Google will not become the #2 player by 2023 is BECAUSE they considered discontinuing the service. They have a history of doing this, as they've axed dozens of other popular Google services that were not as profitable as originally hoped.
Building a business on an infrastructure that might not exist 18 months from now is idiotic.
Re: (Score:2)
Right after Alphabet decided to actually stick with the google cloud offering (for however long they actually do) last year, they bought out Alooma and made it google cloud only. Well, we were depending on going with Alooma for basically all our ETL orchestration and Data Integration orchestration. I won't say who "we" are, but its sort of a big deal to anyone in the United States who cares what kind of government you have and I'm super pissed about it.
Also another side of "we" IS on Google cloud for the
Re: (Score:2)
> Do no evil, my ass.
The amusing thing here is that they even got bored of that:
https://www.independent.co.uk/... [independent.co.uk]
Modern Business (Score:5, Interesting)
I do not get this "I must be number 1 no matter what" mentality. As long as you are making a profit who gives a s**t. There are companies who are # 1 and never made a profit, but the investors are still throwing mega $ at them.
In the old days, if a company made a profit, that was fine. And if you add a new customer once in a while, even better. Now if you are # 1, even if you are running a debt at 200% of your gross and growing, Wall Street is happy
Re: (Score:1)
but the investors are still throwing mega $ at them.
Pretty sure you answered your own question somewhere in there.
Re: (Score:2)
Yep, Alpha Bet has trying new things in its DNA. It's cash cow is Adwords, but it's been trying lots of other things that aren't cash cows. Building better mouse taps doesn't lead to more sales, it only makes better mouse traps, but when your cash cow makes you overflow with money, you try screwy stuff until as mentioned above, you get bored... or the revenue stream is unclear, or competition, or other complications of coffee-born ideas scribbled on napkins.
Re: (Score:2)
The company I work for is well over 100 years old, and operates in an industry that is mature. We made ~$7 billion in gross profit last year, and spent ~$1 billion buying back shares, but some stupid "activist shareholder" still demanded more growth.
Re: (Score:1)
but some stupid "activist shareholder" still demanded more growth.
He should invest in something else.
Re: (Score:2)
It is a whole business model, I am led to believe, but yes, the board very politely told him his plan was a poor one, and he should sell his shares.
Re: (Score:2)
The MBA types who run these places want a 45 degree growth slope for infinity. When that slope tapers off they gut the company of assets and IP and move onto the next.
Re: (Score:1)
The MBA types who run these places want a 45 degree growth slope for infinity. When that slope tapers off they gut the company of assets and IP and move onto the next.
Sometimes they buy up profitable, successful companies and gut them of assets and IP anyway. Especially if the growth wasn't fast enough.
Re: (Score:1)
Corporate governance is really the fundamental flaw in modern capitalism. I suggest we kill all the MBAs and see we get a new, better kind of capitalism. Even if we don't, well, at least we killed all the MBAs.
That's correct for new, growing industry (Score:3)
If you have 5% of the breakfast cereal market and you're profitable, that's great. If you're losing money trying to sell breakfast cereal or soap, you're probably doing it wrong.
In a new industry that is growing fast, the strategy is different. For simple math, let's assume the market in a particular industry is doubling every year. Establishing yourself as the industry leader, with 50% of the market, means you'll have twice as many customers next year. When the market is growing fast, the right play is
Re: (Score:2)
Yeah, I agree. There was a story not too long ago about the last VCR manufacturer calling it quits because they were selling only 750,000 units per year.
Re: (Score:2)
No odder than when a good TV series calls it quits before the inevitable slide to irrelevancy.
Re: (Score:2)
Rapidly evolving industries relies on growth. Due to a number of factors:
1. The tech itself relies on having a lot of users. Anything that involves machine learning and data analytics will naturally mean the guy who's #1 has an upper hand. But beyond that, having high growth means you get to hone your skills at not just making a product but scaling it quickly to an exploding number of users. That in and of itself is an incredibly valuable company asset.
2. Economies of scale. Without enough users, the R&
Um no... (Score:2)
Big investment in GCP (Score:3)
Google did a big rebranding exercise last year and this year has been spending heavily on hiring expensive sales folks , solution architect, pre sales engineers and setting up partnerships with vendor companies to push GCP. They are also presenting on GCP at various conferences. This means they are serious about GCP like they never were about Appengine (GCP prior to rebranding)
Cloud is a commodity. Sure you need to have the tech as table stakes but what really differentiates is how much you are willing to spend on sales.
Google has traditionally not sold to big enterprise. Rather they have setup self service online purchase kind of platforms but cloud investments are big ticket items and you need live sales people to make it happen. Oracle btw excels at this and is also trying to push their own cloud now
Re: (Score:1)
For some reason, Google hired the failed Oracle cloud exec who was fired from Oracle. They hired him to take over the Google clou
Stop the uncertainity, Google, please... (Score:3)
Google has a notorious habit of cutting stuff off. One thing that gets people skittish about GCP is the concern that Google might just announce that they are tired of playing in the cloud arena, are taking their toys and going home.
If Google can show they are committed to the cloud space, they would get more customers. Oh... perhaps some price breaks can't hurt either.
Re: (Score:2)
Executives debated corporate strategy (Score:2)
Uh isn't that what we pay them for?
They should be debating all of it.
Companies that don't discuss and debate their strategies end up missing out. Many great companies have failed or at least suffered greatly from failure to reconsider their business assumptions.
Google's Biggest problem (Score:1)
Hope this doesn't put Kubernetes in danger (Score:2)
I know it's open sourced, but GC depends heavily on it, and so that at least drives feature releases.
Oracle man? (Score:2)
TFA says Google recently replaced the Cloud team's leader with a former Oracle man. Given Oracle's performance in the cloud business, I don't think this bodes well for Google's chances of becoming the second cloud provider worldwide - especially in only four or five years.
Wework and Cloud Computing... (Score:2)
Cloud portability (Score:2)
Nice Post.. (Score:1)
Alternative (Score:2)
Google should get out. They are behind the curve and unlikely to catch up.
Instead they should focus on pushing open cloud standards that make clouds as close to a commodity as possible. That would send the hosting industry to India or Asia*, taking it away from MS, Amazon, etc. and making it cheaper for Google to use.
If you can't beatem', make 'em moot. Google has the money and clout to greatly influence such standards.
* The physical servers may still be semi-local, but most administration could happen in a
Here's advice for google on how to fix their cloud (Score:2)
2. provide good support. no, out sourcing your help to stackoverflow isn't support.
3. fix your damn consoles, they don't make any freakin sense and lack consistency. I've suffered thousand knife cuts with their crappy web consoles
Here is the thing, a lot of people have told them this already. They don't listen and haven't been able to do these little things.