Amazon Web Services Blog: We Build Muck, So You Don't Have To

Earlier today, Amazon.com CEO Jeff Bezos gave the keynote presentation at the MIT Tech Review's Emerging Technologies Conference. The conference was held on the MIT campus in Cambridge, Massachusetts.

There were two distinct parts to Jeff's talk.

In the first part, Jeff made sure that the audience knew the basic financial and numerical facts about Amazon.com - 59 million active customer accounts, 35 retail product categories, 1.08 million seller accounts, and 10.2 million square feet of fulfillment space. Note that these numbers are current as of the date of this blog post, and that you should consult an official source (e.g. Amazon Investor Relations) whenever you need current numbers.

Jeff then made it clear that over the years we had invested about 2 billion dollars in what he called the "guts of Amazon.com" -- multiple data centers, a vast data warehouse, the ability to handle spikey traffic, to deal with fraud issues, and to handle millions of online transactions. He took the opportunity to announce that over 200,000 developers have signed up for our program, and to remind the audience that we now have 10 distinct web services. To wrap up this part of the presentation he presented brief overviews of the Amazon Mechanical Turk, Amazon Simple Storage Service, and the Amazon Elastic Compute Cloud.

To set the stage for the second part, Jeff read verbatim quotes from a number of bloggers including Alex Barnett, Mike Arrington, Bhanoo, and Oliver Thylmann.

In the second part, having set the stage, Jeff proceeded to talk about the changing nature of the obstacles faced by an entrepreneur trying to bring a software product to market. In the shrink-wrapped software business, getting shelf space at physical retailers is a barrier that many wanna-be products never get past. Mediocre products with great distribution regularly outsell great products with mediocre distribution. The web changed the game in a number of ways, removing the limits on shelf space and obsolescing retail distribution as a barrier to success. However, aspiring web entrepreneurs face another battle, namely, doing the undifferentiated yet all-important "heavy lifting" needed to create and operate a web application.

Remarkable aspects of this heavy lifting include server hosting, bandwidth management, contract negotation, scaling and managing physical growth, dealing with the accumulated complexity of heterogeneous hardware, and coordinating the large teams needed to take care of each of these areas. Each area is a career unto itself, doing poorly in an area will cause the entrepreneur to fail, and yet doing a competent job at each one is simply the price of admission. Jeff noted that developers routinely spend 70% of their time doing this backend work. At Amazon we call this ensemble of challenges "muck." Imagine some software developers forcing their way through waist-deep mud, and you'll have a good idea of what he's talking about.

After being in business for nearly 11 years, we know how to make great muck at Amazon!

As part of Jeff's vision for Web-Scale Computing, we are able to share our world-class muck with other developers, giving them the benefit of all that we've learned during those 11 years, and giving them the ability to compete based on the quality of their ideas rather than on their ability to create their own muck.

Important elements of the Web-Scale Computing vision include elastic capacity (scale up or down quickly), high speed, high availabilty, rock-solid reliability, simplicity, cost-effectiveness, and a low up-front investment. I'll have a lot more to say about all of this in some followup posts over the next couple of weeks. The actual products which realize this vision include Amazon S3, Amazon EC2, Amazon Mechanical Turk, and Fulfillment by Amazon.

After his talk, Jeff provided a nice 3-minute summary of his talk on video.

The talk received a considerable amount of online coverage. Here's a partial list: