Amazon CEO Jeff Bezos, photographed at Etech 2005 by James Davidson
Amazon published “disappointing” earnings this week, and the market responded, slashing up to 12 percent off the company’s value. It’s a market that’s preoccupied with fast returns, and so Amazon’s investments in data centers and cloud computing don’t necessarily translate to short-term profits. In other words, the company’s ups and downs highlight contradictions between companies investing to support long-term objectives and running a business to ensure analyst-pleasing growth each quarter.
Amazon has been here before. A BusinessWeek cover story from 2006 captured widely shared sentiment regarding the company’s then-recent foray into building data centers; about Elastic Compute Cloud, the piece found that “Eleven of 27 analysts who follow the company have underperform or sell ratings on the stock — a stunning vote of no confidence.” In September 2011, Forbes painted a different picture, noting, “Amazon shares have defied gravity, jumping 55% year-to-year.”
Those analysts were wrong in 2006. And they’re probably wrong this week. Amazon is not in a high-margin business, where profits come easily. Amazon is dominant in business areas where profits come from achieving massive volumes of extremely low-margin sales. Books, consumer electronics, data center compute cycles: None come close to the 30–40 percent profit margins Apple enjoys.
Nevertheless, Amazon continues to invest in data centers — and to reduce prices — in order to retain its position as the dominant provider of public cloud infrastructure. All of these investments position the company for future growth, with data centers being fundamental to Amazon’s existing cloud business. But all of them hit the bottom line in the latest earnings report, and that has clearly spooked investors once again.
Writing for Forbes before Amazon’s earnings report, CFA Institute‘s Robert Stammers asked whether Amazon’s valuation might boil down to a “cloud bubble,” in which hype and investor exuberance artificially inflate perceived value. Stammers goes on to ask, “Is the cloud a profit opportunity for investors or a product opportunity for companies” (my emphasis). This may be the heart of the problem. Investors are looking for a return that is either large or predictable, and ideally both. Figures such as the $3.2 billion spent by CenturyLink to acquire Savvis make the cloud infrastructure space look like one in which large sums can be made. Maybe they can, but hardware and operations also cost a lot to procure, deploy and maintain. Investors used to software, where profit margins can be 90 percent or more, struggle to adapt.
As Henry Blodget argues at Business Insider, Amazon’s long-term approach to building a business should be welcomed. It is, however, difficult to align with the market’s obsession with short-term growth. Buying into Amazon, or competitors such as Rackspace and CenturyLink, is certainly not a bad idea. You’re not likely to lose money (note: GigaOM Pro is not offering investment advice here). But you’re also not going to see significant returns in the short term.
One of the biggest advantages of cloud computing — for a user — is that providers are able to scale resources in line with demand. But for the user to get that experience, the cloud infrastructure providers (Amazon, Rackspace, etc.) have to spend. Machines have to be bought and maintained in massive data centers, so that they are ready when demand arrives. That spending is expensive, risky and recurring. Amazon has a slight advantage over its rivals because of its scale, and because it can subsidize activities with revenue from other areas of the business. But it still has to upset Wall Street by investing.