Analysts following Amazon.com (NASDAQ:AMZN) rave about the growth of Amazon Web Services. Jillian Mirandi and Michael Barba from Technology Business Research estimate that AWS will generate $3.2 billion in revenues this year. Ben Schachter from Macquarie Capital gives the following estimate for Amazon Web Services:
Using our estimate of $3.8bn for 2013 AWS revenues and applying a ~5x multiple based on the comps noted above, we arrive at a valuation of ~$19bn for the business on an EV/Sales basis (equating to ~$41/share of AMZN stock). Importantly, we believe this to be a conservative valuation multiple, as AWS revenues are growing much faster than any of the comps incorporated above. At an 8x valuation multiple, we estimate the AWS business could be worth $30bn as a stand-alone company, or ~$66/share.
While these estimates may look reasonable compared to Amazon's bubble-level valuation, they do not correctly reflect the value of AWS. Paulo Santos makes a compelling case that AWS margins are close to 10%. I believe that despite its impressive growth in the past, AWS is unlikely to be a profitable business in the future.
AWS faces multiple competitors - most notably, Google (NASDAQ:GOOG), Microsoft (NASDAQ:MSFT) and Salesforce.com (NYSE:CRM). It also competes with the integrated IT vendors like IBM (NYSE:IBM), Hewlett-Packard (NYSE:HPQ) and Oracle (NASDAQ:ORCL), and numerous startups. There is a fierce competition in the cloud computing space. As a result, prices are falling fast.
The biggest problem for all cloud service providers is the relative ease of migration from one platform to another. Customers, especially the larger and more sophisticated ones, are concerned with vendor lock-in - features that tie applications to a particular platform. This creates an unpleasant paradox: in order to attract more customers, platform providers have to stick to open standards and offer most features through the standard APIs, therefore making it easier for customers to move to competing platforms.
Standard APIs also make it possible to mix and match services from different providers based on their prices. For example, App Engine has its own Mail API, but many App Engine customers prefer to use a different service if they send more than a few hundred emails per day. Just a few weeks ago, I decided to use Bing Search API for a new application because it is currently cheaper than Google Search API. If prices change, I can switch to Google's API within a day or two. It is equally easy to swap Amazon's S3 storage service and Google Cloud Storage, or to move applications between Amazon's EC2, Google's Compute Engine, and virtual servers on Rackspace (NYSE:RAX). Obviously, the larger the application, the more time it may take to migrate to a different platform, but customers running large applications have a stronger incentive to move - even a small price differential can save a lot of money.
Consider, for example, monthly data storage prices (per GB) on three popular platforms:
Cloud storage looks like a commodity service, doesn't it? Other cloud services have a more complex pricing structure, but most software developers are smart enough to calculate which service offers the best price/performance ratio for a particular application.
There are customers that are more "sticky" - notably government agencies, public school districts, etc. These are, however, very high maintenance customers. In addition to a lengthy and expensive sales process (including RFPs, proposals, presentations, demos, contract negotiations, certifications, kickoff meetings, progress reports, etc.), they require such a level of customer service that, at the end of the day, they are not very profitable. Often the process repeats itself every 3-5 years. Because the pricing mechanism is very transparent, it is hard to charge these customers a premium for anything other than project management, user training or integration services. Besides, these premium customers rarely buy cloud services directly. Most of the time, they are interested in specific applications. The developers of such applications rip all the benefits - the cloud providers are paid the same price that all the other customers pay.
Commoditization of cloud computing platforms pressures margins for all providers. I see, however, three factors that put AWS in a particularly vulnerable position relative to its competition.
First, AWS is very small. It may be among the leaders in the cloud platform space, but its customer base and sales organization are small in comparison to Microsoft, IBM, HP, Oracle or Google.
Second, the most lucrative part of the market is cloud applications and related services (migration of the existing workloads to the cloud, integration of different services and applications, and outsourcing of the entire functions for large customers). AWS does not have an industry expertise or an army of consultants and project managers to be on the same level with Hewlett-Packard or Accenture. It does not have captive customers like IBM, Oracle or Microsoft. It does not have specialized applications like Microsoft, Google, Salesforce.com, Automatic Data Processing (NASDAQ:ADP), Paychex (NASDAQ:PAYX) or Xerox (NYSE:XRX). In an industry where most players have unique competitive advantages, Amazon has none.
Finally, key players in the cloud space have the entire ecosystems on their side. These ecosystems include operating systems, databases, popular applications, programming languages, development tools, multiple devices, etc.
Imagine a large public school district looking to migrate to the cloud. Most likely, it runs its email on Microsoft Exchange and stores student data using Microsoft SQL Server. All schools in the district use Windows PCs and run Microsoft Office. People working in its IT department spent a decade or more working with Windows, Exchange, SQL Server, and go to Microsoft's conferences every year. Trying to move this district to a non-Microsoft platform is extremely hard - even with a low-ball price.
Google has been recently gaining ground in the cloud space. In 2012, Google announced that 5 million businesses, 66 out of the top 100 U.S. universities, and government agencies in 45 states, adopted Gmail - a cloud-based email service. Now Google can piggyback on Gmail's success and push its Cloud Platform to these customers, using Google Apps, Android tablets and Chromebooks as additional incentives.
Microsoft and Google can (and probably do now) run their cloud platforms at a loss, because they can generate other revenues and because these platforms are strategic assets that solidify their respective ecosystems.
For example, Google Plugin for Eclipse (a very popular tool that developers use to build software) includes both App Engine and Android development kits. It even includes a wizard for creating new Android applications with App Engine as their back end. Similarly, Microsoft's Visual Studio includes Azure tools. When AWS launched, it was a very attractive platform. Today, most developers have to make an effort to even consider AWS instead of Google Cloud Platform or Azure. AWS will certainly continue to have some momentum, because many developers are already familiar with it, but their loyalty will evaporate as soon as AWS becomes more expensive than other platforms.
Google and Microsoft will aggressively invest in and promote their cloud platforms as long as developers use App Engine to host Android (and soon Google Glass) applications or use Azure for Windows/Windows Phone apps, and because they make money on Gmail/Google Apps or Office 365, respectively.
Note that Google Search, Bing, YouTube, Xbox Live, and many other services are using the cloud technologies too. When Google or Microsoft invest in these technologies, they benefit in many different ways, and their R&D dollars are amortized over a much larger revenue base.
AWS has to stand on its own; there are no other revenue streams to fall back on. It does support the other parts of Amazon's empire (like Prime Video), but those businesses are not generating any profits.
Analysts are making a big mistake when they take the past AWS growth and extrapolate it into the future. They make an even bigger mistake when they assume that AWS is very profitable and will remain profitable as it grows. The competitive landscape has changed. The technology has changed. While the number of customers, computing loads and data volumes will continue to grow fast for all cloud platforms, the revenues will grow slower because of the rapidly falling prices. Operating margins for the platforms will be very slim, and it will not surprise me to see negative margins as many leading cloud platform providers prioritize revenues from their applications and ecosystems - something that Amazon cannot do.
Disclosure: I am short AMZN. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.