1
|
They say the future of the web is in web services and if [Amazon][1] is any indicator they're right. Amazon has been working overtime to position itself as the leading provider of web services for a new generation of companies looking to ex
You probably think of Amazon as a store, but Amazon would like to change that impression. Today's Business Week has a long article entitled *[Jeff Bezos' Risky Bet][2]* that focuses on Amazon's attempt to transform from an e-commerce giant to a software company.
One of the big problems with e-commerce packages is that they're either too customized or not customizable enough. Amazon's strength lies in the fact that Amazon, as an article on [Web Services Journal][3] puts it, "eats it's own dog food."
To solve this problem of the gap between what a programmer thinks is necessary in a web service and what a business actually wants, Amazon effectively forked itself into components. As it set about developing web services for internal use, Amazon became a consumer of its own services. The feedback that happened internally at Amazon has led to a package of web services that take some remarkably complex engineering problems and wrap them up in easy to use APIs. That Amazon has turned around and started offering various web service APIs to the world at large, is indicative of it's transformation from online store to online service provider.
And the services that Amazon is now slowly releasing are in fact the same tools it uses to power it's own e-commerce offerings.
From an engineer point of view what makes Amazon's offerings, for instance the S3 storage service, compelling is that the services are completely decoupled from each other. If a small business wants to use S3 to store data but doesn't want any other service from Amazon and in fact wants to use a competing service for the other areas of their business, that's fine. The S3 service has no lock-in with anything else. Amazon has essentially just taken the core principles of Object Oriented Programming and scaled them to encompass whole web services.
While this is a somewhat technical point that may be lost on many observers, for savvy web 2.0 startups this means they aren't entangled in a spider web of interconnected components. It also means Amazon's services scale well. Because of the loose coupling between them, Amazon makes it easy to add new services as a company grows or to discard services that are no longer needed.
It might sound too simple to be remarkable, but Amazon's greatest strength in the web services game may be that it has taken to heart an old engineering quip, do one thing and do it well. In Amazon's case this means have each individual part do one thing and do it well. Doing one thing well is far often more valuable to businesses than a huge unwieldy service that attempts to do everything and ends up doing nothing very well.
But as the Business Week article points out, Wall Street is still scratching it's head trying to figure out where the money is. Analysts seem so far skeptical of a model that takes e-commerce necessities like search, storage, lookup and management of data and turns them into pay-per-use services.
And In some respects Wall Street is right, the target market here is small and medium size businesses, larger companies may well stay with in-house solutions. Is Amazon betting on the long tail effect for revenue? That remains to be seen, but it is worth noting that some big players have already taken advantage of Amazon's services, most notably Microsoft which uses the Elastic Compute Cloud service to help speed software downloads.
Amazon will very likely be rolling out some more services and an overview of it's strategy at next week's Web 2.0 Conference.
Monkey Bites coverage of the Web 2.0 Conference begins on Wednesday, be sure to check here for all the latest news.
[1]: http://Amazon.com/ "Amazon.com"
[2]: http://www.businessweek.com/magazine/content/06_46/b4009001.htm "Business Week"
[3]: http://webservices.sys-con.com/read/262024.htm "Web Services Journal"
|
