Is Jeff Bezos a brilliant idiot or a lucky visionary?
Amazon Web Services are rolling out and getting more play, both pro and con. BusinessWeek online has a recent interview with Bezos, where he outlines what he calls his “developer-facing business” or AWS. The interviewer is dripping skepticism over the project and Bezos doesn’t do much to address his issues.

Yet the premise is simple enough: Amazon offers services on the web. Some are retail services, some are business services and now some are infrastructure services. These services have different business models and the common thread is that if Amazon weren’t building a huge web infrastructure to support the retail business, they wouldn’t know how to get into the other two businesses.

During the California Gold Rush, the people who got rich were the ones selling supplies to the miners. A few miners did well, but most didn’t. Amazon is selling supplies to web miners.

SOA for the people
As I’ve noted (see Efficient Resource Allocation in SOA’s Parts I & II) the big problem that will hold Service Oriented Architectures back in the enterprise is the lack of market discipline. In contrast, Amazon Web Services lives in a fishbowl of market feedback: paying customers, retailers, who watch availability like a hawk; internal developers who use the services; and now external developers.

Virtuous cycle
Building a not-very-profitable scale-out internet data center so one can build and sell scale-out internet data services isn’t an obvious strategy. Yet when a new market is developing there is no better market research.

Amazon, like any business, can be criticized by shareholders for not offering better results, and they frequently are. Yet it takes time build what has never been built for markets that don’t yet exist. If investors don’t want to take that risk they should sell the stock. I, for one, applaud Amazon’s approach and believe it will pay huge dividends in years to come.

Comments welcome, as always. Moderation turned on to defeat comment spam.