Exclusive: The Story of AWS and Andy Jassy’s Trillion Dollar Baby
As the late Stuart Scott would say “AWS has created so much value it’s ridiculous”.
Here is the story of AWS that Dave Vellante and I would like to share with you about AWS and their leader Andy Jassy.
For all of entrepreneurs and developers in the technology business, Amazon Web Services (AWS) is not something new. It has been a godsend in creating value for all. There is no doubt in my mind AWS will go down in history along the side of IBM, Microsoft, Google, Facebook, Hewlett-Packard, and others who created the technology computer business.
AWS, the pioneer in cloud computing, has radically disrupted the global business landscape and the customer experience from the CIA to Apple Computer. Dave Vellante and I sat down with Andy Jassy, who heads up the AWS business, to understand the evolution of AWS and it’s an opportunity to create a trillion dollar brand.
We’re in the middle of the biggest technology chess match in history. The defending champs are the old-guard purveyors of hardware and software: HP, IBM, Oracle, Microsoft, and VMware/EMC. The challengers are cloud computing platforms, pioneered by Amazon Web Services. The stakes are high. If old-guard tech gets checkmated, it stands to lose tens of billions in market value from the traditional hardware and software business models have delivered results for decades.
And it happened almost by luck. According to Andy Jassy who spearheaded and now leads the Amazon Web Services (AWS) business, there was no “ah-ha” moment that lead this online bookseller to one day disrupt a trillion dollar technology market.
Amazon Web Services wasn’t any one person’s single idea. No proverbial apple fell on some Newton’s head, no Henry Ford or Steve Jobs-like character had a brainstorm. Instead, it rather emerged. The idea grew organically out of the company’s frustration with its ability to launch new projects and support customers.
Operating with a focus on delivering better experiences internally and for merchant partners, like Target, to hear Jassy tell it, the concept of AWS was sort of stumbled upon while seeking to solve a recurring need, namely faster technology deployment. This trend would manifest as an entirely new and game-changing approach to technology development, which involved decoupling services. No one at Amazon knew at the time it would morph into what AWS is today.
“When we wrote the business plan for AWS, I don’t think any of us had the audacity to predict that it would become this big this fast,” Jassy recalls. But getting big fast is exactly what happened with AWS and it’s shaking up the entire technology industry.
Less than a decade old, AWS began what has now become a seismic shift in the global technology landscape. Cloud computing — remote computing, or web services, as cloud computing is sometimes called — has been so wildly successful that it’s increased the pace of technology innovation, the rate of technology adoption, and the volume of data in the world by many fold. Today mobile startups can go global in a snap, and although traditional hardware and software giants criticize Amazon as a platform only for startups, AWS now goes toe to toe with old-guard technology every day on capabilities, services, costs and benefits. It’s been radically disruptive to everything about business and the customer experience. Almost every business uses some form of cloud computing today whether it’s through their own software or a supplier’s, and all consumers brush up against the cloud each time they access a mobile app. The symptoms of this disruption are being seen everywhere in the tech business as HP is breaking up, IBM is divesting and retooling businesses and Oracle is moving its revenue stream to the cloud as fast as it can.
AWS has earned the reputation of being the leader in the so-called infrastructure-as-a-service market in a wave that’s permeating every industry sector, from the US government to Apple Computer’s iCloud and other Apple online services. In May of 2013, AWS was awarded an Agency Authority to Operate (ATO) from the U.S. Department of Health and Human Services (HHS) under the Federal Risk and Authorization Management Program (FedRAMP) AWS has since been deemed reliable and secure enough for the Department of Defense, and yes, even the CIA, who announced that the are fully on AWS.
Another massive win for AWS is the Apple’s iCloud business. NYTimes reported earlier this year Apple’s iCloud runs on AWS. An AWS spokesperson declined to comment on the Apple AWS relationship. My sources confirm the NYTimes story that Apple is running at scale on AWS for its’ iCloud and other Apple products.
AWS is the biggest disruptive force in tech history that happened by accident.
The AWS Emergence
A Harvard MBA, Jassy has been with Amazon since the early online bookstore years. It was when he took a unique position as “Jeff Bezos’s Shadow” in the early 2000s (essentially a chief of staff/technical advisor role) that the seeds of AWS would be planted, take root, and then take off.
By this time, Amazon was no longer just an ecommerce bookseller. They’d launched their affiliates business, providing ways for third-party merchants to feature Amazon products, and they’d begun building web infrastructure for other large retail partners. Jassy’s shadowing responsibilities kept him at Bezos’s side daily, and they spent most of their time together surveying the business and asking themselves two questions: What is the business doing well, and what is it not doing well? A series of realizations began to unfold during this process that would hopscotch the business toward the development of AWS in fairly short order.
As Bezos and Jassy dug into and examined their own internal development process and what they were doing for affiliates and partners, they kept bumping into one particular problem. What the business wasn’t doing well was accurately forecasting the time it took to complete projects — both internal and external. Providing a solution for Target, for example, one of Amazon’s early merchant.com deals, was “far more painful than we thought it would be,” Jassy notes.
Way too much time was being spent building the same old infrastructure — storage and database solutions — over and over again. One product development lead after another pointed a finger at infrastructure development as their biggest pain point. Bezos and Jassy had pulled up the rug on Amazon’s development process and discovered a deep thirst for internally scalable, reliable infrastructure services.
Then, recalls Jassy, Amazon began hearing the same things from external partners who knew and understood the space well and were also expressing frustration. First one, then two, then three CEOs declared infrastructure services a top priority and asked Amazon to take a hard look at helping them with data warehousing. It was too expensive, hard to manage, required too much commitment and was riddled with pricey upgrades. It seems in these early days of widespread Internet adoption, a collective grumbling emerged over how much hardware was required to operate at “web scale” — interact globally in real-time with huge user bases. No one had the stomach for the banks of servers they’d have to swallow in order to operate a global online presence.
Along with Google and Yahoo, Amazon was one of the first businesses in the world to operate at web scale, so this was a development process they knew well. When Bezos and Jassy got around to discussing what the business was doing right, building good infrastructure services floated to the top immediately.
“We’d realized in the first ten years we’d built an infrastructure competence deep in the stack — reliable, scalable cost effective data centers to grow the Amazon retail biz the way we needed to,” says Jassy. “But we’d built Amazon so quickly that a number of the pieces of the platform had become entangled.” To develop solutions for external partners, Amazon would need an effective way to communicate with them via hardened APIs, and that meant decoupling the entangled parts of the platform.
“Under the gun to deliver,” Jassy says, “the learnings were magnified.” In the end, the experience did nothing less than change the way Amazon thought about development. They separated data from the presentation layer operating under the theory that their merchant associates would do far more interesting things with the data than Amazon would ever have the time to do itself. And that’s just what happened. Those customers who used the new API increased conversion rates on their web properties by 30%.
Emergence of Intrapreneurshp — The AWS Unicorn Emerges
Amazon AWS team was born as a startup within a large company. Arguably the best “intrapreneurship” venture of our time. Amazon began to look seriously at what the business would look like under a much broader mission: enabling developers build their tech infrastructure on top of an Amazon-built cloud computing plaftorm. They began by asking some key early stage startup questions about the infrastructure services market and their position and opportunity in it:
- Is the space big enough to become a significant part of our business? Yes.
- Is there a market need for a better solution than exists now? Yes.
- Does Amazon have the competency to provide a successful product? Yes.
- Does Amazon have a differentiated approach? Yes — and here it is:
“If you believe developers will build applications from scratch using web services as primitive building blocks, then the operating system becomes the Internet,” says Jassy — an approach to development that had not yet been considered. Amazon asked itself what the key components of that operating system would be, what already existed, and what Amazon could contribute. They realized they could contribute a great deal, because all of the components identified were non-existent.
“We made a pretty quick decision to do it,” says Jassy who wrote the original vision document. He recalls that the AWS OP1 (Amazon-ese for “first version of an operating plan”) made what was then considered an aggressive ask for headcount to get started — a whopping 57 people.
Their first critical decision was whether or not they should build just one service and see how it goes — either a storage solution, a compute solution or a database solution — or build a complete platform with all three from the get go. “All apps need a compute solution, almost all need storage and most need a database,” Jassy says. “Most developers need some combination of the three, so we were strong about the platform approach right away.” In March of 2006, AWS was launched.
More Than A Better Mouse Trap?
“A lot of the services that exist today are useful technologies, but not new,” says Jassy. “There’s still a lot of opportunity to find technologies that do things faster and in closer to real-time.” When it came to AWS, Jassy was right on the money.
The worldwide collective marketplace for infrastructure software services and hardware data center services is in the trillions. Large enterprises historically challenged with managing cumbersome warehouses were chomping at the bit for alternatives. Smaller companies who’d never been able to afford such services were as well. Developers immediately came to AWS in droves. “We were completely taken aback by all the buzz,” Jassy remembers. “We didn’t promote it at all. But thousands of developers flocked to the APIs and used them for all kinds of things that we’d never have guessed they would.”
Soon, developers began asking Amazon to expose other parts of the platform that had nothing to do with data services and AWS answered their calls. “We listen carefully to what developers tell us will help to move their workloads to the cloud so they can spend more time on what differentiates their business,” says Jassy. “There are a few services that almost every application needs, but if you look at the full portfolio of services required, they use lots and lots of other services as well.” Queuing, notifications, search, and content distribution networks are now standard AWS fare — and a key differentiator for Amazon. “Developers want all of it,” says Jassy, and AWS has delivered.
Is the Cloud a Pricing Race to Zero?
The success of AWS has been so disruptive, the following question seems on everyone’s mind today: With an AWS price tag at 20–30% of the equivalent offering from an on-premises supplier (e.g. IBM, HP, EMC), will the trillion dollar Total Available Market (TAM) for technology infrastructure shrink — or is it growing, because easy, cheap, fast ways of storing information beefs up demand? For Jassy, it’s the latter. “Cloud computing enables more use of technology,” he says, “I believe it’s expanding TAM by adding new user segments.”
For many observers the jury is still out. In March of this year, the big three public cloud providers, Google, Amazon and Microsoft, each announced price cuts in quick succession, reminiscent of airlines, which tend to price in packs. At the time, Google called out public cloud providers (presumably referring to AWS) saying that hardware prices had fallen 20–30% over the past five years yet public cloud pricing had only dropped 8%.
Despite the slight, Amazon’s AWS price cuts are as frequent as a politician’s photo opps, having dropped prices forty-two times between 2008 and March of 2014. Critics say that enterprise hardware players cut prices frequently too but fail to market the fact in a transparent manner; hoping to keep individual transaction prices higher in some deals.
The question of industry revenue impact at the macro level remains unclear. Predicting and quantifying TAM for any industry is not easy, and especially hard for technology because it’s innovating at break-neck speed. The array of technology solutions and use cases changes daily in today’s environment. Pinning it down to a simple quantified number is a near impossible task.
Predictably, Jassy would prefer to take the conversation away from competitive price wars and focus on customers. Almost every business thinks of it’s own addressable market in roughly two buckets: small business and the enterprise. “We knew that the largest consumers of infrastructure would be large enterprise because they spend more absolute dollars,” Jassy says, “But we also had a mental image of a college kid in his dorm room having the same access, the same scalability and same infrastructure costs as the largest businesses in the world.”
By all accounts, Amazon has delivered on that vision as its customers include small startups and large, established organizations in virtually every industry. AWS has a compelling value proposition for each case, and they’re actually quite similar. Large companies are saying yes to AWS because it’s simpler, less costly and less headache, “but what they really care about is agility,” Jassy says. “The most compelling part of our value prop is being able to innovate more easily and quickly. Engineering teams and LOB leaders have a lot of ideas and a strong appetite for doing new things.”
Amazon’s offer to do the “non-differentiated heavy lifting” has enormous appeal to big companies because it frees up more time for their engineers to innovate. Jassy claims that ceding the more rote aspects of tech development over to AWS actually changes enterprise team dynamics. Where product development teams used to get demoralized over the 6–20 weeks it took to get a server so they could experiment, “they’re able to try ideas they’ve had for years in minutes,” Jassy says. With AWS, enterprise developers can “spin out” thousands of servers in minutes. If their experiments fail, they can give them right back. ‘It stirs up the creative juices of the team because they know they can actually get their ideas out there.”
“Invention”, says Jassy, “requires two things: 1. The ability to try a lot of experiments, and 2. not having to live with the collateral damage of failed experiments.” Satisfying both criteria, cloud computing has been a perfect breeding ground for invention. In other words, if developers were pastry chefs, they can now stop spending most of their time milling flour and churning butter, and start spending most of their time combining and presenting those ingredients in new and different ways. And with very little surcharge for failure. Flour and butter is now readily available and inexpensive. It also brings the barrier to entry way down — a huge appeal for startups and large corporate enterprises. Up and coming pastry chefs can get in the game easily, pump out a few never-before-seen delicacies, and test the appetites of the entire world. So what ends up happening? The world becomes awash with intricate, nuanced, original pastries the likes of which no one has ever seen. Just take a look at your iPhone and the enormous amount of applications changing the world.
Today every imaginable workload sits on top of AWS — media distribution, high performance computing, batch analytics, disaster recovery, archival and back up, CRM, ERP, right down to marketing and advertising. While AWS was at first liberating for the enterprise, it’s now an imperative. With all that technology infrastructure floating above the world and there for the taking, deep competitive advantage awaits any business that can innovate fast in this market.
Competition is not happy. Roman Stanek, CEO and founder of GoodData, took AWS and Jassy to the “woodshed” in his open letter to Mr. Jassy, Tear Down This Wall. Stanek was referring to AWS competitive dominance saying “the way AWS behaves needs to change. You built the leading infrastructure-as-a-service provider with a level of secrecy typical of a stealth startup or a dominant enterprise software platform vendor. It works for Apple — they deliver a complete integrated value chain. But it is not your position in the cloud ecosystem”.
Amazon’s enterprise infrastructure competitors are also responding to the “cloudification” trend. For example, Oracle, whose CEO a few years ago made jokes about cloud is now all in, citing fast growth rates in the company’s cloud business. Similarly, Hewlett Packard has formed an entire division dedicated to moving customers to the cloud, while IBM last year purchased Softlayer for $2B in an effort to compete more effectively for cloud opportunities after losing a cloud deal to Amazon at the CIA. Notably, Amazon’s enterprise IT competitors all emphasize so-called hybrid cloud, a form of cloud computing that combines on-premises computing with that of external cloud resources. Jassy claims this is because the hybrid model doesn’t disrupt the legacy company’s cash flow and profits. Traditional enterprise players fire back that Amazon doesn’t understand the requirements of hardened enterprise IT practitioners.
There are other signs that the Amazon AWS formula is gaining momentum. Andreessen Horowitz, a leading Silicon Valley venture capitalist, struggles to understand how Amazon makes money saying “On one hand, there is the ruthless, relentless, ferociously efficient company that’s building the Sears Roebuck of the 21st Century. But on the other, there is the fact that almost 20 years after it was launched, it has yet to report a meaningful profit. AWS is a black box and a much debated puzzle, but it is also pretty much the definition of a new business that requires investment to grow.”
Is AWS Recession Proof?
Born out of the ashes of the dot.com bust in the early 2000s, Amazon also weathered the 2008 downturn well. “The 2008–2009 economy was rough and we didn’t know what that would do to this fast-growing AWS business,” Jassy remembers. AWS had huge adoption among startups, but in that rough economy, many were not getting funded. “But a lot did and I can’t help but to think that AWS was a part of that,” says Jassy. AWS enabled innovators to prove the validity of their ideas in the marketplace, mitigating a lot of risk for investors.
“The recession actually accelerated the rate at which enterprises began looking at AWS as well,” Jassy continues. “CIOs and CFOs were hard pressed to lower variable expenses and were looking hard at every project. I think the recession advanced enterprise consideration of the cloud by at least two years.” When enterprise did come knocking, Jassy remembers the whole team was thinking “This is happening faster than we thought — we should staff up.”
But it wasn’t done in frenzied rush like so many high-growth tech companies are famous for doing flush with lots of cash from venture capitalists. AWS was careful in its’ early expansion period. “We think about our hiring a lot,” says Jassy. “Businesses often compromise on hiring characteristics in the name of rapid growth, but we’re vigilant about hiring builders — inventive, entrepreneurial, creative types that want to operate what they build,” he explains, ”We want missionaries, not mercenaries — people focused on building businesses that last beyond their tenure at the company.”
In typical market-driven fashion, they then fleshed out the AWS business according to the feedback they got from customers. “Some customers never talk to us and don’t want to. Others really want people services,” Jassy says. “The enterprise requires more human contact — dedicated account managers that know their business and their workload as if they were part of their team. Other customers have voiced the need for help thinking through their architecture and building migration plans, so we’ve added solutions architects.”
When it comes to product development, Jassy cites the same market-driven approach. When customers said that for archival process, they’d trade latency for lower prices, AWS built Glacier, an archival and back up storage service at a penny per gigabyte per month.
The Future of the Cloud
So what about the chess match? Will there ever be a clear winner? Will server racks and datacenters in fact go the way of dinosaurs and dodo birds?
“There’s substantial migration happening now,” says Jassy, and it’s delicate work. It requires thought and care, attention to legacies and dependencies that have built up over time, and can take 2 to 4 years to complete. “But repeatable migration patterns can be built,” he adds, noting that all 1000 of Unilever’s websites sit on AWS, and that 500 of them were migrated.
“Our point of view is that in the future very few companies will own their own data centers, and those still in operation will cover a small footprint,” Jassy says. “The vast majority of workloads will go to the cloud. We’re just at the beginning — there’s so much more to happen, especially in the mobile space,” he adds, “Most people believe mobile devices will one day become the primary computing shell. The smaller the device, the more the cloud is needed.” Eventually, Jassy sees on-premise workloads dwindling to only those that involve regulatory or compliance components, or have some other cultural reason for not being migrated. Traditional enterprise IT players see things differently. Joe Tucci, EMC’s Chairman, for example, has often stated that he believes there will be very few public clouds and many private or hybrid clouds in the future. Can both camps be right or is this a war of wordsmithing?
Currently, AWS does a lot of business in its version of “hybrid or migrated workloads” — apps built on premise but that want to reach out to the cloud to supplement. NASDAQ’s Market Watch application, for instance, was built and still runs on-premises, but uses AWS for all their analytics and data crunching, moving data in and out of the cloud several times per day. Nokia mobile apps use AWS Red Shift analytics because according to Jassy it’s half the price and twice as fast as their old on-premises solution. Samsung SmartHub runs on AWS, but their payment transactions are on-premises. According to Jassy, customers seamlessly move between the two without ever knowing it.
The pace of innovation at AWS continues to run very high. “In the sixteen years I’ve been here, I’ve noticed how easy it is to focus on the shiny new pennies,” says Jassy, alluding to the natural tendency for Amazon culture to get excited about new capabilities. He then tosses out a list of AWS services with the enthusiasm of someone describing their first trip around the galaxy. “Elastic Compute Cloud, Elastic Map Reduce, S3 which stores trillions of objects, Dynamo DB — a very fast, high throughput, low latency, non-relational database, RDS Service, Mobile Push Notification — demand for these services just keeps accelerating. We’re waiting for it to attenuate and it just hasn’t yet.”
Where’s all the follow-on competition?
As wildly successful as AWS as been, there still exists no other credible challenger in public cloud with revenue nearly as large as AWS. “We always believed that there would be multiple successful players,” says Jassy, but admits they’ve yet to arrive on the scene in a big way. One of his theories is the market just isn’t ready for the new world order. “AWS has pursued an extremely disruptive, lower margin model in a high-margin industry. It’s conceivable that there are those out there who still hope cloud computing doesn’t get traction.”
And what of the wildly unpredictable open source — the Linuxs and MySQLs of the world, which AWS both consumes and purveys — how does that fit in? “We view open source as a companion to our business model,” says Jassy, “Our customers want to run services on top of open protocols because it gives them the freedom to move workloads around. We use it, contribute to it, think it’s important to the development community and it’s not going away any time soon. So long as our customers are asking for services based on open source technologies, we’ll continue to add them.”
The View from the Top
AWS has pioneered a technology infrastructure that trades capital for variable expense. It has cut the cost of doing business down to the lowest it’s ever been, provides real elasticity, the ability to move quickly, and lets innovators spend their scarce resources on driving product and service differentiation, not housekeeping. You might think that looks like a pretty good cake all by itself, but wait — here comes the icing. Jassy and the AWS team’s competitive strategy is to be faster at iterating on new features to push new innovations at amazing speed. Why? Because it’s all in the cloud, upgrades are free.
This culture of constant iteration means they add capabilities very fast. Amazon Red Shift, a data warehouse service lunched in February of this year has added 20 features in 8 months — without upgrade or maintenance fees and headaches. Incidentally, the Red Shift product has ponied up the fastest growing revenue stream in AWS history. “I think there are some companies that either fear innovation entirely or just do it through acquisition,” says Jassy. “At Amazon, we love innovation. We’re builders here.”
Interestingly, when asked what keeps him up at night, Andy Jassy doesn’t have much to say at all. After a long, long pause, he tells the following story.
In mid-1997, the rumors began to flutter. Barnes and Noble was about to launch a website — bad news for Amazon, still then making all their money selling books online. A company-wide meeting was called — easily pulled together quickly for what was then just another startup of around 200 employees. In that meeting, Andy Jassy remembers a remark Jeff Bezos made that would end up becoming the cultural north for this modest little book retailer destined to disrupt global technology.
“Don’t lose sleep fearing your competitors,” Jassy recalls Bezos saying. “If you’re going to lose sleep, loose it over fearing whether or not you’re delivering the right experience for your customers.” From that simple declaration, the seeds of AWS were already being sown — a technology that would one day call the global technology marketplace to account, a simple byproduct of one company’s mission to deliver great experiences. Today AWS customers include NASA, the Obama Campaign, Pinterest, Netflix, Apple, and the CIA. While AWS revenue is not broken out in Amazon financials (listed in the “other” category), in 2014, industry watchers estimated it to be over $4 billion.
Authors Note: My business parter and Co-CEO of SiliconANGLE Media Inc Dave Vellante (@dvellante) who is also chief analyst at Wikbon Research and cohost with me for our @theCUBE video program contributed to this article. Dave and I sat down with Andy Jassy last year for inside look at how AWS started and grew over the years.
Recently many have been asking about our media business SiliconANGLE Media. We have grown over the past 5 year with three brands covering all the angles of cloud, infrastructure, mobile and social for enterprises. The three brands are: SiliconANGLE.com, Wikibon Research, @theCUBE. We have a free wiki for research at wikibon.org