According to analysts at Morgan Stanley, Amazon’s cloud computing division could be a shining star (even if not too bright) on the company’s long road toward increased profit margins. However, while their forecast isn’t glowing, it also doesn’t account for the evolution of Amazon’s cloud business from pure infrastructure-as-a-service into higher-level (and higher-margin) services.
In the new research report published Monday morning, analysts Scott Devitt, Andrew Ruud and Nishant Verma come to a possibly disconcerting conclusion for any investors banking on Amazon for short-term returns. The report’s gist: “After analyzing Amazon.com’s cost structure in detail and by segment, we conclude that there are far more variable costs than investors believe, leading to an overly optimistic timeline for margin expansion.” But Amazon Web Services is an opportunity Amazon might be able to exploit.
The report estimates that AWS was responsible for $1.19 billion in revenue in 2011 (I predicted in October the business was on a billion-dollar run rate), of which $108 million (or about 9 percent) was sheer profit. It’s able to maintain this margin while constantly dropping prices on its cloud services, the report contends, because AWS uses a cost-plus pricing model. That is, it just adds a premium (about 10 percent) on top of the cost of delivering those services, which continue to drop as Amazon leverages its economies of scale to buy and operate more gear and bandwidth at lower prices.
Although AWS margins remain flat, the report notes that AWS also comprises a significant portion of Amazon’s overall technology spending, so being able to drive steady, predictable profit from it is a good thing. Non-AWS technology spending, the authors estimate, is about 4 percent of net sales — “represent[ing] the largest opportunity for operating margin expansion in the near-term.” Keeping those cost down means a greater percentage of revenue goes toward profit.
However, the Morgan Stanley report doesn’t address the possibility that AWS margins actually could start rising as the company expands its services beyond sheer infrastructure and into managed services. Its NoSQL DynamoDB database service, for example, is a service for which Amazon adds value (and cost) beyond just the delivery of cloud-based infrastructure, and there are lingering rumors of a big data analytics service that will provide higher-level services than AWS’s existing Elastic MapReduce offering.
We shouldn’t overlook the possibility of AWS expanding its licensing activities, either. As it becomes more entrenched as the de facto cloud computing platforms for many companies, providers of other services and software are keen to get on board. Already, private-cloud pioneer startup Eucalyptus has licensed the AWS API, and Citrix wants to do the same for its CloudStack software. If it’s feeling greedy, Amazon could look to capitalize even further by charging others to integrate directly with its business.
Or it could just give that cost-plus dial about a quarter turn.
Image courtesy of Flickr user comedy_nose.