Amazon's Cloud Dominance

Investment Analysis Clubs Saul’s Investing Discussions
1

Pretty good article on Seeking Alpha b Arne Alsin on the AWS cloud dominance. Alsin points out that slower tech companies like IBM, Oracle, and HPQ almost inevitably lost while Microsfot and Google have done well…so far. Nothing new there I know. But this author makes the case that the cloud will eventually be a zero sum game and AWS will be the eventual winner with Microsoft and Google eventually losing too.

I’m not convinced of that last part, but found the article was well worth reading nevertheless.

Five years after Jeff Bezos and Andy Jassy launched Amazon (NASDAQ:AMZN) Web Services, in March 2011, Businessweek reporter Ashlee Vance penned an article titled “The Cloud: Battle of the Tech Titans.”

This article foretold the eventual obsolescence of the mainframe server business and the impending dominance of Amazon in the nascent cloud industry.

Why am I bringing up this five-year-old article now? Because I believe Vance’s predictions are finally coming true.

…

The problem for many of the companies trying to compete in the cloud is this: Cloud, in my view, is a zero-sum game. As Amazon continues to land big name clients and invest in its infrastructure, I believe it will simply pull further and further ahead of its competitors in the space, who will - Microsoft and Google included - lose out in the end.

Research firm Canalys now estimates that spending on cloud revenue could grow to $190 billion by 2020. And it’s my belief that Amazon is in the perfect spot to capture much of that market.

Definitely take the time to read the whole thing if you get a chance: http://seekingalpha.com/article/4031799-amazon-killing-compe…

Matt
Long AMZN, GOOGL, MSFT
MasterCard (MA), Nestle (NSRGY), PayPal (PYPL), and Verizon (VZ) Ticker Guide
See all my holdings at http://my.fool.com/profile/CMFCochrane/info.aspx

16 Likes
2

Of course, for every bull case, there’s a bear case. I include this here only as a counterbalance.

First, Amazon cannot sustain such magical growth. Sooner or later, the margins will even out. Right now, Amazon is gaining market share with AWS by cutting prices and constantly adding services. But when growth levels out, Amazon will be forced to operate on value. Amazon has no experience with value, which implies profitability. Heck, even at its best, Amazon is not very profitable. Its operating margin in 2015 was a meager 2.1%. Faced with the need for profitability, Amazon would probably begin raising prices to cloud-based customers. The dazzle of extraordinary growth will diminish; investors will become disaffected and look around for a more thrilling, explosive company.

Second, Microsoft (NASDAQ:MSFT) is sneaking up on AWS. The manner in which Microsoft reports revenues from its cloud services makes it difficult to know precisely where the company stands compared to AWS. Simply put, Microsoft combines the revenues from software, infrastructure and cloud into what is designated as commercial cloud. This reporting method tends to conceal what’s really happening, while simultaneously bloating the numbers. But by dividing the number by twelve, we can come up with a ball park figure that is fairly accurate. For the first quarter of 2016, which concluded at the end of September, Microsoft’s yearly commercial cloud revenue topped $13 billion, which breaks down to $1.083 billion, or slightly over $3 billion per quarter. During the third quarter, which also concluded at the end of September, AWS indicated revenue of $3.23 billion.

Since Microsoft’s number includes SaaS revenue, it is clear that AWS remains dominant. On the other hand, the fact that Microsoft offers SaaS, along with IaaS, means Microsoft’s cloud-based services have more to offer. According to GeekWire, a survey by Morgan Stanley predicted that Microsoft’s Azure will dominate IaaS and PaaS by 2019, surpassing AWS.

There’s more bearish points made. You can read more at http://seekingalpha.com/article/4031732-invest-amazon-bad-ti…

Matt
Long AMZN, MSFT
MasterCard (MA), Nestle (NSRGY), PayPal (PYPL), and Verizon (VZ) Ticker Guide
See all my holdings at http://my.fool.com/profile/CMFCochrane/info.aspx

5 Likes
3

Matt,

I like your bull case better.

Frank,

:slight_smile:

2 Likes
4

I read the bear case last night and this was my comment:

>>>You’re In For A Bad Time Over The Long Term

That’s what Lord Keynes said. “In the long run we are all dead.” The fun part comes before that.

http://seekingalpha.com/article/4031732-invest-amazon-bad-ti…

The bull case rubbed me wrong, the statement “Cloud is winner-take-all, zero-sum game” tells me the author does not know the right terminology which means he might not really understand the issues. This might sound like hair spitting but the real issues that need to be understood are the reasons for Amazon’s dominance.

The statement “Cloud is winner-take-all, zero-sum game” should be replaced by “Cloud is an increasing returns business, the more you sell the more you sell.” But I’m not convinced that this last statement is entirely true either.

When you have two (or more) incompatible competing technologies the successful one drives the others out of the market or catches well over 50% of the market as a minimum. A typical case was VHS vs. Beta (which many of you might be too young to know about) two different formats of video cassettes which could not be played in the other equipment. Take telephony, there are several standards that coexist. Qualcomm owns CDMA which competes against but coexists with GSM.

CDMA vs. GSM: What’s the Difference?
BY SASCHA SEGAN
FEBRUARY 6, 2015 10:03AM EST246

Two basic technologies in mobile phones, CDMA and GSM represent a gap you can’t cross. They’re the reason you can’t use many AT&T phones on Verizon’s network and vice versa. But what does CDMA vs. GSM really mean for you?

http://www.pcmag.com/article2/0,2817,2407896,00.asp

But CDMA networks talk to GSM networks! While they are incompatible they coexist and that is the case with Cloud. The client browser does not care what standard the server uses.

What this adds up to is that Cloud is a commodity but which benefits from the network effect. AWS’s main advantage, as I see it, was its early start, the so called “path dependence.” Everyone else is playing catchup which gives AWS a huge but not unsurmountable advantage.

Yesterday I made an interesting discovery thanks to a Fool directing me to a database service:

A really interesting product in this area is DataDirect Cloud which allows connecting to a large number of different database types, including NoSQL, but presenting a local interface which behaves like SQL so that one can access it with familiar analytic tools.
https://www.progress.com/cloud-data-integration

http://discussion.fool.com/a-really-interesting-product-in-this-…

What caught my eye was:

We host the solution in AWS and manage all the updates necessary. All you need to do is access the service.

https://www.progress.com/cloud-data-integration

In conclusion, the age of the Gorilla Game domination is over but the network effect is still very much alive. To take an extreme bull or bear position on Cloud providers is not a good idea but clearly AWS is dominant which is a huge advantage.

Denny Schlesinger

14 Likes
5

Another comment I added to the Amazon bear case:

DCF works OK for steady state and slow growth companies – if you guess the inputs right – but DCF is a bad advisor for investors in fast growth companies because Mr. Market pays for growth which is not contemplated in DCF calculations. The risk, of course, is that growth will eventually slow and the P/E ratio will contract but that does not happen overnight. There is plenty of time for the observant investor to take profits and to exit the position as maturity approaches.

http://seekingalpha.com/article/4031732-invest-amazon-bad-ti…

Denny Schlesinger

6

Denny,

Thank you for sharing your thoughts. Another area where incompatible systems co-exist is Windows vs. MacOS.

Cheers,
Chris

7

As I understand the advantage AWS has, it’s a virtuous circle as described below:

AWS has an advantage in having lots more cloud computer centers and lots more services that it provides. This brings in more customers, so they have lots more cloud revenue than their competitors. More cloud revenue means they can build more cloud computer centers and develop more cloud services. And that’s the virtuous circle. Besides, adding additional customers almost costs zero, so you can cut your prices per customer, and still make more money.

To massively oversimplify, if you are bringing in $XXX in revenue, and building a new computing center costs ½ of $XXX, you can only build two new computing centers at most before your are putting out your own money. If you are bringing in 10 times $XXX in revenue, you can build eighteen, and have plenty of money left over for R&D into new services.

This means that AWS can keep bringing in net profit while Google and Microsoft probably have negative margins. Google and Microsoft have plenty of cash, so they can keep pouring money into it if they want to, and maybe, maybe, catch up some time in the vague, undefined future. It’s a bit like smart phones, where Google has lots of phones, but Apple has all the profit.

But that’s just my naïve, oversimplified way of looking at it. I’d love to hear comments, either pro or con, on my thoughts.

Saul

3 Likes
8

This means that AWS can keep bringing in net profit while Google and Microsoft probably have negative margins. Google and Microsoft have plenty of cash, so they can keep pouring money into it if they want to, and maybe, maybe, catch up some time in the vague, undefined future. It’s a bit like smart phones, where Google has lots of phones, but Apple has all the profit.

Saul,

One of the things that I love about you and your approach is that you’re able to cut through the hype and get to the bottom line. Ultimately, a company, at some point, needs to produce profits.

Cheers,
Chris

2 Likes
9

This means that AWS can keep bringing in net profit while Google and Microsoft probably have negative margins.

I don’t get the negative margin thinking, why can’t the other big boys make a profit just like Amazon.
If Amazon is making good money at it they’re not giving their space away cutting profits to the bone.

JT

1 Like
10

AWS has an advantage in having lots more cloud computer centers and lots more services that it provides. This brings in more customers, so they have lots more cloud revenue than their competitors. More cloud revenue means they can build more cloud computer centers and develop more cloud services. And that’s the virtuous circle. Besides, adding additional customers almost costs zero, so you can cut your prices per customer, and still make more money.

Amazon got into this position by starting out first, call it “first mover advantage” or “path dependence.” Catching up is difficult because Amazon has built the network effect where “the more you sell, the more you sell.” Google has this advantage in search, Microsoft has this advantage in Windows and Office, but that does not translate into advantage in Cloud. Amazon conquered a virgin field. Google, Microsoft, et al, have to compete with a well established provider.

But Cloud is still essentially a commodity where the low cost provider has the advantage. That’s why Amazon keeps cutting costs and reducing price. Reinventing The Bazaar by John McMillan as a very good explanation that goes more or less like this:

Suppose there are two competitors making identical widgets. A has a cost of $8 per unit while B has a cost of $10 per unit. At a street price below $10 B goes broke. At a street price of $10 A makes a $2 profit while B makes no profit at all.

As long as Amazon is the low cost producer it can make a profit at prices that drive the competition broke. Google has lots of cheap servers as well but probably not designed for Cloud computing. When it comes to the actual technical details, I don’t have a clue.

Denny Schlesinger

Reinventing the Bazaar: A Natural History of Markets by John McMillan

https://www.amazon.com/Reinventing-Bazaar-Natural-History-Ma…

You purchased this item on May 2, 2005.
View this order

Isn’t it nice of Amazon to remember my details? Wanna bet that these data are in the Cloud?

4 Likes
11

As I understand the advantage AWS has, it’s a virtuous circle

Saul

Kinda-sorta. I see a lot of analysis spouted about Amazon and AWS that ignores some of the fundamentals about why & how they got into the business and why they stay.

As a technology company, Amazon focusses on developing the technology that it needs to operate its ‘regular’ business. Once it has figured out how to do this, it looks for opportunities to leverage that technology with other customers. AWS came out of a need to solve the (server) resource-surge that it’s retail business experienced periodically around holiday periods. Once it hits the spot with customers for the technology, it will continue to develop more independently, but fundamentally it is still a technology that Amazon can and does leverage itself, so economic improvements - due to scale or whatever - will also impact other parts of Amazon’s business.

This symbiotic relationship between ‘core’ business and cloud provision also exists with someone like Google, who uses a sh1t-load of technology in its own right across its various businesses. Rather than trying to sell the ‘raw’ technology, Google has followed the path of wrapping it around other services such as storage (Google Drive). The same is true to a lesser extent with AAPL in terms of their cloud offering, although they probably have a much lower organic need for the underlying technology…their activity appears more defensive, not wanting to cede aspects of the user relationship (storage principally) to others.

So with these companies, there is direct, core benefit from leveraging cloud technology as a business. What about MSFT and ORCL then? The offerings here are more akin to AMZN (raw technology) than GOOGL (services packaged around the technology): they want customers to use their virtual servers for work rather than buying their own or renting from a niche hosting company like Rackspace, and they have less potential for leveraging what they learn in other parts of their business. However, MSFT and ORCL each have a key advantage when looking at cloud provision to the business segment: they are able to provide customers with very easy access to machines running with their database software, which is a vital software component for many companies.

MSFT seems to have realized the importance of this a few years ago and created some special licenses for folk like Rackspace to use on their boxes - they wanted to make it more likely that smaller companies would use MSQL rather than the open source MySQL. They have shown signs of changing their license architecture for MSQL, which might enable them to give Azure an advantage, but this has not yet filtered through to the marketplace. However, they have made Azure VERY easy to implement and use. (I’m sure AWS have made huge strides, but I remember the hoops one had to jump through back in 2008 to spin up a box!). They need to think about how to organize their ‘menu’ of machines, but I’m sure they will get there. Companies that I am associated with still prefer using Rackspace and AWS (for different applications), but we did use Azure for a couple of projects last summer just to try it and it was not bad at all. And stupid cheap. Certainly much improved from 2 years ago when we looked at it last and gave up.

I have no direct experience with ORCL installations, but I suspect that there is a bit of a mis-match between the mentality of customers who want expense-based technology facilities for not much cash, and what ORCL is used to charging for its database software. Their model may well end up being something different: high-priced virtual hosting for premium customers.

Hope this helps

Cham

18 Likes
12

I agree AWS is and always will be the largest winner in Cloud.

However I reject the idea that it will eclipse all others. (IMHO)

The reason for this is

1, IT hates monopolies. We learned a lot by dealing with Microsoft. Even now ‘How to’ and ‘Getting started’ stories on Cloud computing all (I have not read every one) include a caution that you should have access to at least two Cloud systems, and your code should be written to be portable from one to the other.

2, Amazon is a known entity when it comes to pricing. They will always charge the most they can. This means that if the market ever let them have a monopoly, even if just perceived, then they will raise prices. It won’t be all at once, and it won’t be on every product, but prices will go up, until someone else says “Hey I can do that and charge less”

2 above feeds 1. 1 helps hedge 2.

3 Likes
13

I know that it’s commonly accepted to say that Amazon had first mover advantage, but I think that saying this obscures an important point.

Long before the advent of ‘the cloud’, many companies were pursuing the same abstract dream of capturing part of the trillion dollar IT infrastructure market by outsourcing it to their own distributed network platforms. AKAM in 1998 just to name one of a hundred, ranging from small startups to tech giants like CSCO. The problem for all the companies trying was how to envision and frame the offering, how to market it, and how to capture critical mass of participants. It’s not so much WHAT Amazon did as it is HOW they did it. As I said, many were trying. We don’t know most of their names because they failed to capture the critical mass that Amazon was able to.

It’s more like they have ‘first succeeder’ advantage. Yes, they were the first to move toward what is now seen as the right answer, but they also invented the right answer.

2 Likes
14

I retired from a Fortune 50 company in 2010. Prior to retirement I was an Enterprise Architect. At the time, “cloud” was just beginning to be a thing. While I’ve not tried to keep up with the industry or technology (duh, I’m retired). Yet, I still understand the basic concepts, and I have friends (one of whom is Chief Cloud Architect) for the company I used to work for. Here’s my take.

Cloud really started out as nothing but remote storage. It solved a lot of problems. Companies saw their storage server bill growing on an annual basis. The company I worked for sold products with extremely long life cycles. Many of their products would remain in service for 50 years or more. They provide spare parts and emergency repair services for every product until the last one is retired. Because of this they had both legal and business reasons to maintain certain documentation for inordinately long time periods. This further led to an ethos within the company to unnecessarily retain a lot of documentation long after its utility expired and it was really a liability.

Then, with the explosion of the internet and more and more business activities migrating from paper to digital medium, storage demands skyrocketed. If all the retained information had been recorded on paper, the company would have had to figure out where to put it. They used Iron Mountain for storage of stale documents, ones unlikely to ever be looked at again. But it wasn’t digital. Because of the peculiar way IT charge-backs worked, managers did not directly budget and pay for the storage used by their units. Storage was free and invisible. IT had to budget and plan for virtually all the information storage for the entire company.

Enterprise storage is not trivial. Low-latency access requirements must be met. A lot information must be available 24x7x365 world-wide. Routine back-ups have to be created and retained. Disaster recovery must be provided. And of course, security is a major issue. And those are just the big, easily understood requirements. Litigation creates a whole set of additional legal requirements to support discovery. M&A activities generate a whole new set of requirements, and so forth.

When it comes to IaaS, cloud is basically a commodity. But, SaaS is a different critter. From what I’ve read, and from what my former colleagues tell me, MSFT is pulling ahead of AWS and others in easy to use cloud software and service offerings especially geared for the enterprise customer.

And another thing almost universally over-looked by the financial analysts is that MSFT has had an enterprise presence since the days of DOS on an 8086 PC. WindowsNT quickly became the dominant internal network s/w. MS-Office is in every major enterprise in the world. MSFT has a long, deep presence in the Enterprise environment. Amazon does not. There is immeasurable value in the personal relationships between MSFT business and technical interactions and support and enterprise customers.

So long as the customer is seeking only to reduce the complexity and cost of digital storage, a lot of cloud providers can address the requirements. But, when value added functionality becomes part of the requirements, MSFT is in a much better position than AMZN.

MSFT began as a company that unseated the industry leader (Digital Research had majority share with the CP/M operating system in the tiny but rapidly growing micro-computer market). MS-Word unseated WordStar and a host of other word processors. Excel, completely overtook 1-2-3 and Lotus, Access quickly dominated the PC database market. PowerPoint displaced Harvard Graphics. In other words, MSFT has repeatedly beat out other companies with first to market and dominant share advantage. It would not surprise me if MSFT repeated this success with Azure.

Given that, I’m long AMZN and not MSFT. AMZN is much more than a provider of AWS. MSFT does not come close to covering the waterfront that AMZN does. But if you focus strictly on cloud, it would not surprise me to see Azure surpass AWS in the not too distant future.

18 Likes
15

2, Amazon is a known entity when it comes to pricing. They will always charge the most they can.

I wonder where you get that idea. I was subscribed to AWS in the early days when I was still developing websites. Every so often Amazon would lower their price, again and again.

I’ve been with my current web host during 13 years, they raised price once about 10 years ago (40%). Since then the price has held steady but the service has grown and grown, more memory, more bandwidth, offsite backups, offsite emergency server, faster servers, more redundancy.

The market simply won’t allow price gouging, it’s too competitive. Bezos would have to be an ass to price himself out of the market. Bezos is not known to be an ass. BTW, Amazon would also be stupid to charge less than the market will bear.

As for IT hating monopolies, it hates job insecurity even more. That was a great help to me while I was an IBM salesman. It worked against me while I was selling NCR. “You can’t lose you job for hiring IBM” was the mantra back then. I somehow doubt that the new IT crew is more altruistic than the old. LOL

Denny Schlesinger

16

I built my own company in 1978. Sold it in 1998. My customers were
Tech hardware corporations. I was a precision machining and plastic
injection molding corp. Short run protypes and longer run production.

We took any job that was feasable and we had very high quality standards
In the beginning I thought I had to do lower pricing in order to keep getting the jobs.
Eventually it dawned on me that these guys really valued
what we were delivering to them. So I started raising my prices.

I pushed the bubble and was surprised at just how high I could go with
higher pricing. In return I gave them the best they could get. Usually
better than most of the local competition.

Profits are very important Without profit I could not afford to stay in
business. Machine tools, injection molding machines, facilties,
other manufacturing technologies of the time, payroll etc etc etc.

I charged, for my services, however much the customer was willing to pay.

If customers are willing to pay for Amazon cloud services, whatever it is that they are charging,
then Amazon will have no reason to lower their prices. And so it goes, for any business.

Frank

4 Likes
17

Frank makes a good point. I have never understood the term ‘price-gouging’. The beauty of capitalism is that, in a free market, the seller will naturally charge whatever the market will bear while competition works on behalf of the consumer in eroding that advantage.

2 Likes
18

I wonder where you get that idea.

Amazon has a reputation for testing pricing all the time. I recall people at work pulling up the same item, and getting different prices, based on their viewing history, and likely many other things we were not aware of.

I would guess that they would do something similar here. I believe they still publish prices for AWS for individuals and small businesses, but not for enterprise level support.

The market simply won’t allow price gouging, it’s too competitive.

That was my counter point to the idea that all Cloud computing will end up at Amazon.

As for IT hating monopolies, it hates job insecurity even more. That was a great help to me while I was an IBM salesman. It worked against me while I was selling NCR. “You can’t lose you job for hiring IBM” was the mantra back then.

It was a mantra, but not a universally held truth. You took that job at NCR and I am certain you did not lose every contract that included IBM. Also no one says that about IBM anymore.

Groups first starting out will still tend to use AWS, but once they are there and up and running for a while, the sales people from the other systems will show up, and some of them will get business just for the sake of diversity. I work for a large corporation and we use AWS and at least one other Cloud service, but we also maintain our own clusters.

Mark.

3 Likes
19

Mark, what’s important for an investor to figure out is whether AWS does or does not have network effect working in its favor. If it does it’s a great investment, if it doesn’t it will be a drag on Amazon. So far the numbers, the share of market, say that AWS is king of cloud.

I had a similar discussion as far back as 1999 about ARM Holdings, the upstart that was taking on INTEL but going after the low end of the market. People were saying that INTEL would wipe them out. Seventeen years later ARM customers are shipping ARM based cores by over a billion monthly. It was a question of looking at the numbers

Some people missed the ARM tornado

The last remaining question is the one about the tornado. Some people are waiting for the tornado. In my opinion a two year long tornado ended a year ago with the happy result that ARM garnered 80% of the RISC core market:

 Quarter   Revenue   Growth q/q
 6/30/98    16,568    
 9/30/98    19,022    
12/31/98    20,753    
 3/31/99    21,110    
 6/30/99    22,787      37.54
 9/30/99    25,850      35.90
**12/31/99    30,482      46.88**
 **3/31/00    30,115      42.66**
 **6/30/00    34,995      53.57**
 **9/30/00    39,028      50.98**
**12/31/00    44,560      46.18**
 **3/31/01    46,211      53.45**
 **6/30/01    50,619      44.65**
 **9/30/01    59,074      51.36**
12/31/01    56,925      27.75
 3/31/02    60,006      29.85
 6/30/02    65,794      29.98
[http://discussion.fool.com/some-people-missed-the-arm-tornado-18...](http://discussion.fool.com/some-people-missed-the-arm-tornado-18511025.aspx)

During all these years people would show up saying that INTEL would eat ARM’s lunch. It hasn’t happened. On the contrary, ARM is threatening INTELs bread and butter, the server market. Opinions don’t matter. Look at the numbers.

Denny Schlesinger

Chips containing ARM processors shipped per year
https://www.theatlas.com/charts/Ek18VmbP

5 Likes