All Shook Up – How AWS is Pulling Off the Impossible in Enterprise IT


December 4, 2015

Summary:

  • Amazon Web Services (AWS) is growing like gangbusters and taking enterprise IT by storm, quickly becoming the safe choice for corporate CIOs.

  • Against all expectations, the cloud infrastructure service provider is throwing off profits while it leaves its competitors in the dust.

  • How did the company achieve so much in such a short time, and what does this mean for the traditional global IT players?


“The Tipping Point is that magic moment when an idea, trend, or social behavior crosses a threshold, tips, and spreads like wildfire.”

– Malcolm Gladwell, author of the bestselling book “The Tipping Point”


Since around 2008 AWS, as both thought and action leader of the public cloud infrastructure ‘movement’, has been very successful with tech startups and small to medium businesses in many industries. Now, against steep odds, Amazon’s cloud infrastructure-as-a-service division appears to have recently passed a critical tipping point in the enterprise, becoming the clear number one choice for boards, management teams, and CIOs. That said, it is still relatively early in the cloud business as far as larger corporate and government organizations are concerned, so we shouldn’t get too far ahead of ourselves.

But let’s consider some of AWS’s accomplishments to date. According to industry analyst firm Gartner, this year AWS, is delivering over 10X the utilized cloud capacity of its largest 14 competitors combined. Deutsche Bank analysts Karl Keirstead and Ross Sandler reported in a note dated November 3 that in its most recent quarter, AWS grew 78% year over year, reporting $2.1 billion in revenue, after growing another 81% in the quarter before that. Furthermore, AWS had profits of $521 million last quarter, giving it a healthy 25% margin. Based on its growth, Deutsche Bank says AWS could be the fastest-growing enterprise tech company ever. As the analysts also stated: “AWS launched in 2006 and is forecast to surpass the $10 billion revenue mark on its 10th anniversary next year. By comparison, it took nine years for Google to hit $10 billion and 10 years for Facebook. Oracle and Microsoft each took 23 and 22 years, respectively, to pass the $10 billion milestone. … Measured by revenues, AWS is approximately 6x larger than its biggest rival Microsoft Azure and is arguably the greatest disruptive force in the entire enterprise technology market today. … We conclude that a 2017 revenue multiple of 10x seems fair for AWS, which given our 2017 AWS revenue estimate of $16 billion implies a valuation of $160 billion.”

On this last point, I believe there’s a case to make that AWS is already worth almost as much as Oracle, which has a market cap of $160 billion, and that it is worth triple the value of Salesforce, founded eight years before AWS was launched and by far the largest cloud application (Saas) software maker, with a $50 billion market cap. The reason for this bullish assessment of AWS’s ballooning valuation is that AWS’s 70%+ annual growth – which vastly outstrips the growth rate of the mother ship, though admittedly on a much smaller revenue base – when combined with the power that AWS now exerts in the marketplace, is to my mind responsible for most of the 115% run-up in Amazon’s stock price in 2015 and all of the 75% run-up since the stock was at $385 two years ago today (Dec. 1, 2013). Such is the magnetic field and stratospheric valuation of an emerging gorilla franchise. Despite the fact that we are still in the relatively early days of cloud computing in enterprise organizations, this lead is indeed beginning to look insurmountable.

As a result, traditional incumbents such as HP, IBM, Cisco, Dell, and EMC (the last two currently trying to execute a $67bn. mega-merger) are all aquiver about how to manage the decline of their traditional businesses and get onto the cloud in a meaningful way. They’ve wasted too much time second-guessing the speed at which large business and government organizations would overcome their earlier concerns with data security and opt for having (some of) their systems hosted in the public cloud versus other hosting alternatives such as private cloud and managed hosting. In particular, they never expected that an upstart like AWS could win out against more established enterprise vendors like Microsoft or IBM.

Indeed, AWS’s only legitimate competitor today in Iaas and Paas is Microsoft, currently a distant though fast-growing #2. With its Azure infrastructure and platform assets, Microsoft is expected to be able to build a strong business around its Windows, Office, ERP and CRM applications, and .Net offerings which are now being delivered as subscription services.

Before moving on to try to explain what AWS itself has done to become so successful in the enterprise in such a relatively short time, we should note a few critical developments in the past two decades since cloud computing first became a reality for those of us who started sending email, browsing web sites, and using e-commerce in the mid-90s. At that time we didn’t use the term ‘cloud computing’; instead, we were “using the internet”, or “web-surfing”, or “shopping online”. But these were the early precursors of Saas applications. Three developments were, I believe, instrumental in establishing a fertile adoption environment for cloud infrastructure services that AWS, Google, Microsoft, IBM others are now tapping into.


Three Critical Developments in Enterprise IT

First, over the past decade or so enterprise organizations have been “softened up” with respect to adopting the cloud by the enormous success of Salesforce, whose SFA (Sales Force Automation) application took full advantage of the target-rich environment left by the kludgy, integration-heavy client/server CRM implementations of the 90s and early 00s. Salesforce.com the application was an iteration of cloud infrastructure-as-a-service in action, proving itself to be sufficiently reliable and secure, and much more compelling to use, than its client/server predecessors. For the first time ever, salespeople, managers, marketers, product managers, finance managers, legal counsel, and others clamored to get access to this tool that gave them direct line of sight into customer opportunities as well as customers’ evolving requirements, usage and ongoing engagement.

Secondly, with numerous other multi-tenant applications emerging in the past dozen years from companies like Concur, Eloqua, Financial Force, Hubspot, Iconixx, Intacct, Magento, Marketo, Netsuite, ServiceNow, Successfactors, Workday, and Zuora, business users and their IT staff began to appreciate the ease and flexibility of subscribing to these services in the cloud. Over time this growth and increasing prevalence of Saas adoption put AWS, the clear thought and action leader in Iaas, firmly in the forefront as the favored choice for providing the underlying computing infrastructure for thousands of B2B tech startups, and later departments and divisions or larger organizations in other industries.

Thirdly, it was critical that the middle layer in the cloud stack – Paas (platform-as-a-service), crucial for enabling customers to develop their own applications when off-the-shelf Saas apps didn’t fulfill all their needs – achieve greater adoption, so that the three-layer cloud stack could be seen by customers as a reliable architecture to invest in for production systems. Paas had lagged Saas and Iaas for some time, slowing down the growing cloud juggernaut. Customers were leery of adopting one vendor’s platform service versus that of other vendors, partly to avoid lock-in and partly because of considerable confusion about the make-up of the Paas category itself. Initially, the variety of offerings from different Saas and Iaas vendors such as AWS’s Elastic Beanstalk, Google’s App Engine, Salesforce’s separate platform’s Heroku and Salesforce1, Red Hat’s Openshift, Pivotal’s Cloud Foundry, and IBM’s Bluemix and other platforms made customers skittish about potential lock-in to a proprietary platform that did not feel ready to commit to. Recently, however, Microsoft’s Azure has gained increasing acceptance as the casing that houses all of Microsoft’s franchise offerings including Office, its ERP and CRM business management applications, and .Net development tools. Furthermore, Salesforce now provides offerings that overlap into the platform area such as Saas extensions, development services, and APIs, and AWS now provides application containers, development tools, and application lifecycle management tools through their own efforts as well as through their ever-enlarging partner ecosystems. With these recent evolutions and Microsoft’s surge into second place behind AWS, customers are now more confident of placing bets on applications hosted in the cloud because they see a legitimate player to keep AWS honest and a more consistent architectural cohesiveness in the entire stack.


How AWS Pulled Off One of the Greatest Feats in Enterprise IT History

So how has AWS done it? First, let us not forget that Amazon has its own well-proven “iteration” (or Saas application) of its infrastructure service, which it spent more than a decade refining before even launching AWS – i.e. Amazon.com, the e-commerce business that so many of us use today for online purchases of everything from books to baby products. What better way to hone your own infrastructure (and platform) capabilities than using it/them as the foundation for running your original core business? On top of this, unlike any of its competitors except Google (currently a very distant #3) with its search franchise, Amazon has been able to leverage the e-commerce business’s massive investments in datacenters around the globe, and in essence fund the development of AWS out of the company’s e-commerce operations. In other words, AWS did not exactly start at zero (that said, Microsoft Azure is funded out of Microsoft’s enormous cash reserves, as is Google’s Cloud service out of its search franchise).

This point about leverage leads directly to the second development: While Amazon’s e-commerce business has routinely reported losses or at best minuscule profits, AWS is in reality – as reported by the Deutsche Bank analysts and others – highly profitable, partly as a result of operating at much larger scale than any competitor by a mile, partly through its mastery of datacenter design and provisioning, and user experience design, and partly through its rapid and relentless pace of innovation. The business model is based on the original insight that Jeff Bezos and his team had about marketing the spare processing capacity that the e-commerce business would not use in off-peak periods and that otherwise would only weigh on the company’s balance sheet. Thus, AWS can afford to use price reductions as a weapon against Microsoft and Google whenever it chooses to and it will still throw off greater profits than its competitors. This is how Amazon fooled everyone – including myself – who thought that they were waging a “race to the bottom” (how many times have we heard this phrase in connection to Amazon in recent years!).

The third, and possibly most surprising, development has been the speed with which AWS has migrated from serving almost entirely cloud-friendly startups and small businesses to serving highly symbolic enterprise-grade organizations such as the CIA (after winning the deal “twice” in a harshly disputed competition against IBM) and being declared by GE’s CIO in a recent conference as GE’s cloud infrastructure provider for the future. Other major corporations such as Comcast, J&J, Kellogg’s, Pfizer, Qantas, Nordstrom, and Unilever have also made strong commitments to the AWS service. What’s happened during the past two or three years is what always happens when a company becomes the de facto gorilla in a major category: they become a magnet for customers and partners and as long as they don’t become too hard to deal with, a powerful partner ecosystem forms around them – be they Saas vendors companies, platform or tools providers, consulting firms, systems integrators, ad agencies, or others. This force-field helps to explain why financial analysts like Deutsche Bank start foreseeing major upticks in the company’s valuation. Partners are able to plug the gaps that AWS’s business model (based on operational excellence in delivering highly automated, high-volume transactions) would not otherwise fill. This is not to say that AWS is unable to deal with enterprise customers. In fact, they are building a considerable enterprise sales and support capability. However, service partners such as CapGemini, Cognizant, Unisys and others, provide relationship management and systems migration and integration coverage for enterprise customers.

Overall, Jeff Bezos, Andy Jassy (AWS CEO), and their team have executed an impressive strategy based on the customer journey illustrated below.

A New Gorilla Rattles the Enterprise IT Cage

I mentioned earlier my belief that AWS is firmly ensconced as the new gorilla in enterprise IT. More importantly, it is the dominant leader of a revolutionary shift in how organizations, not just individual users, consume IT. It’s worth keeping in mind the definition of a bona-fide gorilla franchise. This phenomenon occurs (a) when one company becomes perceived as the dominant player in an important high-growth category, (b) when the perceived switching cost of changing to another provider is a palpable deterrent, and (c) when it becomes clear that the gorilla can really only lose their dominant position if they screw up through malfeasance or incompetence – no other condition is sufficient to cause their fall from #1. Such has been the case of Google in ad-based search, Oracle in relational databases, Microsoft in Windows and Office, SAP in client/server ERP, Cisco in network switches and routers, and IBM in mainframe computers. Multi-billion dollar franchises, all. Whereas previously many observers (again, including myself) would have argued that cloud infrastructure was rapidly becoming commoditized, and that there would be insufficient switching costs to deter customers from moving from one provider to another, I now believe that the ecosystem that has converged around AWS, which adds platform and even Saas capabilities on top of the infrastructure service to form a more compelling set of offerings, makes AWS the newest gorilla in enterprise IT. As I intimated in a recent post, I would not be at all surprised if in 5 years’ time we look at Amazon as a cloud infrastructure services giant that has a flourishing retail “application” on the side.

Thus it is clear to me that AWS has surpassed the major tipping point of becoming the safe choice for mainstream customers in cloud infrastructure. Its focus, agility, and innovativeness have put it years ahead of its closest competitors, with a much more extensive and ever-enlarging partner ecosystem. This is the power of gorillas: they don’t have to be the easiest to deal with – and AWS is reputed to be difficult to deal with at times – but they concentrate so much power in the market that partners gravitate to them in droves.

The only other gorilla today in the cloud world is Salesforce.com, in Saas. It too has the type of power that I’ve described above as applying to AWS, having been the original thought and action leader behind the “end of software” movement. What is interesting is that what has taken Salesforce.com 17 years to achieve, having been the stalwart application software pioneer in cloud, AWS has achieved in just over half the time. When you put AWS and Salesforce together, you have arguably a combined $200bn. valuation. While by no means a measure of a mature business, this is definitely a mark in the evolution of the cloud as a mainstream offering for pragmatist customers and makes it increasingly appealing to more conservative-minded buyers.


So, What’s a Shaken-Up Competitor to Do?

This leads me to the last section of this article: What can or should AWS’s competitors in Iaas and Paas do to compete effectively for a good share of the pie going forward? Can any company give AWS a run for its money, or is it game over? What about Microsoft and Google, supposedly the most dangerous threats to AWS’s hegemony? And, even more important in some respects, how will AWS’s breakout success in public cloud impact the major enterprise computing incumbents such as IBM, HP, Dell, EMC, and Cisco, whose growth has slowed and whose relevance, to one degree or other, has regressed in recent years?

It’s been very good to see how Satya Nadella has come in and dramatically reset priorities at Microsoft. In part because he ran the emerging Azure business before becoming CEO, he has been responsible for getting the company to harness its various offerings into a “cloud first” portfolio including the massive Office 365 franchise. .Net, etc. Although Windows-friendly IT operations may well choose to go with Azure rather than AWS, the portion of the world that follows Microsoft is smaller than it once was and this is unlikely to change. In other words, while AWS is cool, Microsoft is still seen as kinda square. Thus Microsoft will most likely be the lead chimp forever in this category, albeit most probably a quite successful and profitable one.

The search giant Google is a different case. To the surprise of many who have foreseen Google catching and eventually overtaking AWS, it is now a very distant third, especially as a choice for corporate and government customers. There seem to have been three or four spurts by Google in marketing their cloud services, but they still don’t seem fully committed to the cause. Furthermore, they don’t seem to have an enterprise bone in their body, and don’t look like growing (or acquiring) one. It’s just not who they are. So while startups everywhere use Gmail, Google Docs, Calendaring, and other applications, I can’t see enterprises ever choosing Google’s Cloud services over AWS or Azure in any significant numbers, just as I don’t see Google staffing up with marketing, sales, and technical resources to support them, or building a sufficiently strong partner ecosystem to do so. In fact, Google reminds me strongly of Apple in its aversion to the B2B market.

As for the traditional paid-up members of the enterprise IT club, the biggest test of their ability to adapt to the new consumption model for enterprise IT is going to be one of culture and business model. The shift from mainframe to client/server caused the downfall of major companies such as DEC, Data General, McCormick & Dodge and many others. In contrast, the shift from the on-premise perpetual license culture and business model to the consumption model is far more challenging for incumbents to deal with. Sales forces can no longer sell and leave, but nor can they get accustomed to the demands of customers who can leave at any time; product development is no longer upgrade-driven, but requires a more dynamic dev/ops mindset; and support is no longer about break/fix, instead requiring a focus on user adoption and continued engagement. Virtually all of the senior management ranks in these companies grew up executing to this now outmoded model. It’s no less than a generational shift, seismic in its impact. That’s why I believe, and have said in prior blogs, that the old-guard companies will most likely need to pull off reverse mergers in which younger acquired companies provides many if not most of the new leaders. But this dramatic set of changes is unlikely to take place without its share of shakes and tremors, as we’re seeing with the struggles that IBM, HP, and Dell are experiencing.