Amazon Web Services, which provides computing and storage services to hundreds of businesses, began as a seemingly crazy idea in 2006, but it has since grown into a $1.5 billion business this year, according to a new report from R.W. Baird. It's believed to be one of Amazon.com's fastest-growing businesses, the largest piece of an "other" revenue category that grew 68% in the third quarter, to $648 million, far outpacing overall company revenue growth of 27%.
Today, Amazon founder and CEO Jeff Bezos offered an explanation of how AWS, often seen as something that has little to do with its core retail operation, fits into Amazon's business, and how it runs on similar principles. In fact, he says, it's quite similar to Amazon's Kindle business, where the company makes little money on the device but a lot more if it's used--in that case to buy lots of books and movies.
Bezos made his comments, which were webcast early Thursday afternoon, to close the company's first public conference on Amazon Web Services, a three-day geekfest that started Tuesday in Las Vegas. Following are his edited and sometimes paraphrased comments in conversation with Werner Vogels, Amazon's chief technology officer, sometimes ranging well beyond AWS to entrepreneurship, rockets, and a humongous clock in the mountains of west Texas. (Unfortunately, he had nothing to say about Amazon's surprisingly large ad business.) You can view the whole keynote here or click the video above.
Vogels: The last time you were onstage, at the Kindle Fire announcement, you said that Amazon should only win when our customers win, and that's how the Kindle Fire business works. I'd like to think AWS also works that way, but elaborate on that.
Bezos: It's very similar to our Kindle device business. We sell our hardware near break-even, so we make money when people USE the device, not when they BUY the device. That is very aligning with customers. It causes us to have the right behaviors.
AWS is very similar. It's a pay-as-you-go service. We're not incented to get people to overpay for hardware. In the long run, that will work out very well for customers, and it will work out very well for Amazon.
Vogels: You've always talked about flywheels, which in Amazon retail is low prices, convenience, etc. What's the flywheel in AWS?
Bezos: I always get the question, what's going to change in 1o years? I almost never get asked, what's NOT going to change in the next 10 years? That's the more important question, because you can build a business around things that are stable. One is low prices. Another is faster delivery. Vast selection.
On AWS, the big ideas are also pretty straightforward. People will want more reliability, lower prices, etc. The big ideas in business are often very obvious. But it's hard to keep a firm grasp of the obvious.
Vogels: What are the real mechanisms of innovation?
Bezos: Innovation is a point of view. You have to select people that want to innovate, to explore. An explorer company isn't for everybody. But for people who get up in the morning and want to change things, it's a lot of fun.
Other things important for innovation isn't as fun. One is you have to have a willingness to fail, to be misunderstood for long periods of time. Then, you can ramp up your rate of experimentation. Successful invention requires you to increase your rate of experimentation. AWS is one of those things that helps startups do experimentation faster.
Vogels: What's been most surprising about AWS?
Bezos: It's the growth rate of AWS over the last six or seven years, especially with enterprises, education, and governments.
Vogels: How important is access to near-perfect information that you've talked about to Amazon's business?
Bezos: I see this as a societal trend that has been going on for a long time, accelerated by the Internet. A lot of friction to do comparison shopping has been reduced. The balance of power is shifting from the providers of services to the consumers of services.
If you have a business model that in part relies upon your customers being misinformed or incompletely informed, you better think about changing your business model.
Vogels: There's a feeling of "lean manufacturing" in AWS and customer service.
Bezos: One of the key things about "lean" is that you need a lot of information, to keep defects from moving downstream. ... People who are fans of lean manufacturing take almost as an article of faith that the further a defect travels downstream, the more expensive it is to fix.
All the people at Amazon do customer service training sessions. I got a particular call once where the customer service person predicted that the person would want to return the table she ordered. And she said she wanted to return the table. I asked how he knew she would want it returned. He said, that always gets returned, because of a packaging defect that puts a scratch on the top. So now a customer service person can pull off a product if they see a problem. We're not going to ship the same table to the same customer over and over.
We can apply all these ideas to Amazon Web Services.
Vogels: How do you operate a low-margin business like AWS?
Bezos: Operating a low-margin business is hard. I sometimes have waking dreams that someday I might operate a high-margin business. It's impossible to be efficient if you have high margins because you don't need to be. We wouldn't know any other way. It keeps you honest.
Vogels: How has entrepreneurship changed in the 18 years since you started Amazon?
Bezos: Some of the basic principles haven't changed. But in the Internet arena, one things that's changed is that the rate of change is even faster. The heart of entrepreneurship is risk-taking, but good entrepreneurs seek to reduce risk. When a company gets big and robust, you can start taking risks again.
Vogels: What is AWS's role here?
Bezos: We're still in the early stages of seeing what AWS can do for entrepreneurs, especially inside big companies.
Vogels: Can AWS be a business that can go after enterprises at the same time as startups?
Bezos: Yes, because we're already doing it. We have an existence proof. Everybody wants the same things. Everyone wants low cost, reliability and innovative services.
Vogels: Should we be worried that AWS serves both Amazon Instant Video and a competitor, Netflix?
Bezos: No. We treat them just as well. We bust our butts every day for Netflix. Amazon retail gets the benefit of AWS standardizing these services.
Vogels: Moving to something completely different, tell us about the 10,000-year clock.
Bezos: I am building along with other people, including the father of the idea, Danny Hillis, in a mountain in West Texas a very large clock designed to last 10,000 years. It's a way of ensuring long-term thinking. If we think long-term, we can accomplish things we couldn't otherwise accomplish. You can't solve world hunger in five years. You can do it in 100 years. You create the conditions for that to happen.
Vogels: What are you doing in space? What's the state of Blue Origin?
Bezos: It's a company developing a vertical-takeoff, vertical-landing spacecraft. It's designed to be reusable. We're on our third development vehicle. I'm hopeful this third one will be our last before we build our operational vehicle, take people to suborbital, sell tickets.
Every approach has its pluses and minuses. You need reusability for low cost. You need to practice, and you can't practice with expendable vehicles. It's a bad cost structure. So you have to get the vehicle back. One way is with wings. Another is to land it vertically under rocket power. One thing I like about that is that it scales better to larger sizes than winged vehicles.
Blue Origin is a huge user of AWS.
Vogels: What would you suggest to people who want to start a company?
Bezos: Never chase the hot thing. [Though one can argue that's precisely what he did, judging from his own words about starting Amazon because he saw Internet traffic skyrocketing.] You need to position yourself and wait for the wave. To do that, it needs to be something you're passionate about. Missionaries build better products. Mercenaries want to flip the company and get rich.
Number two, start with the customer and work backwards.
Those two things will take you an awfully long way.
