While there is a wide range of opinions on the long-term ceiling for Amazon.com's (AMZN -0.89%) retail business, the company's cloud-computing arm is an even bigger wildcard. Amazon Web Services allows other companies to rent computing power from Amazon rather than buying and operating their own servers.
This business has been growing very rapidly. Last year, revenue in Amazon's "other" segment in North America -- which primarily consists of AWS sales -- reached $3.7 billion: up from less than $1 billion in 2010. The rapid growth of Amazon Web Services has led some analysts to value this oft-overlooked part of Amazon's business at $50 billion or more.
To justify these estimates, analysts typically point to other "cloud-computing" companies like salesforce.com (CRM 3.59%), which has a $34 billion market cap despite generating just over $4 billion in revenue last year. But Amazon Web Services is nothing like Salesforce. It is a much more capital-intensive business than Salesforce and has minimal pricing power.
Another round of price cuts
Amazon Web Services (like Salesforce) is the clear leader in its field. In fact, AWS dominates the "infrastructure-as-a-service" market. According to Gartner (IT 1.23%) it is more than five times the size of the next 14 competitors combined! Normally, this kind of market share lead should translate to significant pricing power.
But AWS is under pressure -- or at least feels under pressure -- from various Big Tech firms that are trying to break into this market, including Google (GOOGL -1.33%) and Microsoft (MSFT -0.68%). Neither company's IaaS offerings are close to AWS in market share, but Google and Microsoft both have unique competitive advantages in cloud computing. Google has its popular Google Drive service, various productivity apps, and Compute Engine, while Microsoft has its Office 365 cloud-based productivity suite as well as Azure.
In any case, Amazon has been ultrasensitive to the prospect of being underpriced by other IaaS providers. Since Google launched its AWS competitor in 2012, the two companies have periodically engaged in price-cutting battles.
This past week, Google cut its "App Engine" prices yet again. It didn't take long for Amazon to shoot back with price cuts of its own that range from 36%-65%, effective April 1. AWS senior vice president Andy Jassy made light of the price cuts, noting that AWS has reduced prices 42 times in its relatively brief history.
Jassy is essentially trying to make a virtue of a clear weakness. If AWS needs to respond in kind to every competitor's price cuts, it will have two negative impacts on the business. First, it will cause revenue growth to slow dramatically -- if prices fall 50%, AWS would need to double its volume just to keep revenue flat. Second, it will ensure that AWS remains a perpetually low-margin business.
A black hole of CapEx
The commoditized nature of Amazon's product stands in stark contrast to Salesforce's business. But an even bigger difference between the two companies is the capital intensity of their respective businesses.
Unlike AWS, which primarily sells infrastructure as a service, Salesforce offers software as a service. While Salesforce has to devote a significant amount of money to R&D -- over $600 million last year -- the vast majority of its costs relate to selling, general, and administrative expenses.
SG&A has been rising rapidly recently, but over time Salesforce should be able to leverage those expenses significantly. This potential for margin growth can justify Salesforce's high valuation even though it has consistently lost money on a GAAP basis and has a narrow profit margin on a non-GAAP basis (which excludes acquisition-related costs and stock compensation costs).
By contrast, as AWS grows, it needs more and more computing power and storage to rent out to its clients. This has driven a massive jump in Amazon's capital expenditures recently. From 2005 to 2009, Amazon's annual CapEx rose from $204 million to $373 million, which was significantly slower than the rate of revenue growth.
By 2013, CapEx had skyrocketed to $3.44 billion: up nine-fold in just four years. Some of that growth is related to new distribution centers for the retail business, but a large portion can be attributed to AWS.
For Salesforce, strong revenue growth brings the promise of margin expansion as the marginal cost of adding customers is fairly low. That's not true for Amazon Web Services. As it gains more business, it must plow billions and billions of dollars into things like servers, storage, and networking gear to support that growth.
Foolish bottom line
Amazon Web Services should be a solid contributor to Amazon's overall business in the long run. But just because it's involved in "cloud computing" doesn't make it similar to companies like Salesforce.
Despite its dominant market share, AWS competes tooth and nail with competitors like Google and Microsoft. Moreover, it needs to spend an ever-increasing amount on adding computing power in order to support its growth. This makes it a fundamentally low-margin, weak-cash-flow business: not one worth a gaudy multiple like Salesforce.