Speaking at the University of Washington, Andy Jassy, head of Amazon Web Services, said that luck played a part in AWS’ huge growth — but early decisions paved the way for the company’s cloud-computing dominance.

Share story

The head of Amazon Web Services (AWS) said its dominance in cloud computing stems, in part, from luck — but also from early decisions that proved critical to establishing a revolutionary business.

“There’s always a fair amount that is luck,” AWS CEO Andy Jassy said Tuesday in a speech at the University of Washington’s department of computer science and engineering. “You have to have the right timing and some things have to break your way.”

Jassy’s comments come as AWS — Amazon’s cloud-computing unit, which rents out computing power and storage to enterprises, governments and entrepreneurs — has become a $14 billion-a-year business, driven in recent years by mass migration of data from companies’ private data centers to shared ones. It’s a decade-old business that Amazon pioneered — and a hugely profitable one.

In 2016 it brought home $3.1 billion in operating income — 32 percent more than Amazon’s North America retail unit, the company’s largest business by revenue. It’s also, by far, the largest cloud provider.

AWS’ sales growth rate, however, has slowed to 47 percent in the fourth quarter of 2016 from 69 percent in the year-earlier period, amid stiffening competition from Microsoft, Google and other large tech firms.

Among the early decisions at the foundation of AWS’ fortunes was the creation of “primitive building blocks,” basic functions that customers could use and combine according to their needs, Jassy said.

Also key: AWS sold its services a la carte and based charges on usage, much like a public utility. That was a big departure from the expensive, multiyear contracts that technology providers typically charged. “People gave us a lot of credit early on for the pricing model,” Jassy said.

The next critical decision was the market AWS first went after.

AWS’ leaders “very consciously targeted software developers and startups early on,” even though they knew that enterprises and governments would eventually be the largest clients, according to Jassy.

“That turned out to be an extremely underserved segment,” Jassy said. A lot of those developers were spending only a few bucks on AWS services, but “we didn’t mind that,” the executive said. “Some of those are going to be the next big enterprise in the next five to 10 years.”

It was also necessary to keep innovating, quickly, to adapt to growing needs. “Being quick and moving fast and being feature-poor to start with only works if you can deliver and iterate quickly,” Jassy said.

Now the breadth and sophistication of AWS’ offerings include artificial intelligence, voice computing, databases and machine learning tools, some of which draw on the innovations Amazon has deployed on its Alexa digital assistant and in its fulfillment centers.

As for the growing competition, Jassy said that the market for computing-related services is so big that there is room for a number of successful cloud providers. But “I don’t think there’s going to be 30, because scale really matters.”