While Amazon dominates cloud infrastructure, an oligopoly is emerging. Which will buyers bet on?

Kurt Marko Profile picture for user kmarko February 15, 2017
Amazon may hold a large lead over Microsoft and Google but that doesn't mean Amazon is invincible. The market is large enough to allow for the creation of a oligopoly. And that's exactly what we see emerging.

By inventing the market for public cloud services, AWS enjoyed a significant first-mover advantage over competitors that ceded Amazon years to develop services, build infrastructure and woo customers. But Amazon is only part of an emerging oligopoly where customers will have real choice.

The result is that, regardless of how you measure it, whether in service maturity, market share or revenue, AWS maintains a significant lead in the public cloud market.

So far, the cloud behemoth has done nothing to jeopardize its dominant position. Its R&D engine is still in high gear as evidenced by over 300 new services and features introduced last quarter in conjunction with a re:Invent conference that drew record crowds of over 80,000 live and streaming participants, all while continuing to cut prices.

It's a good thing AWS isn't slacking off because there's a big, fat bullseye on its back.

If nature abhors a vacuum, business abhors a hegemon, where competition makes it increasingly difficult to maintain massive market domination while the accumulation of past financial success makes it impossible to sustain astronomical growth rates: ask Apple.

While AWS no longer doubles in size every year, its growth remains impressive. However there are subtle signs that it could be starting to saturate the pool of available customers, if not their usage. And it is happening  just at the time other cloud services emerge as viable alternatives.

The result will likely be a public cloud market that resembles other capital-intensive business segments like autos, semiconductors and integrated energy companies that are dominated by an oligopoly of large, sophisticated players, accompanied by specialist providers. You can almost see that in the sense that SAP is not regarded as a public cloud market play yet it is building out capability aimed at running its largest and most demanding customers. That's not something AWS aspires towards any time soon.

Amazon's fourth-quarter results might have disappointed Wall Street, with the stock down nearly 5% in the days following the report, but the cause certainly wasn't AWS. Profits in Amazon's cloud segment accounted for nearly three-quarters of the company's operating income last quarter. Indeed, the stellar growth at AWS might explain why the stock has almost entirely recovered from the sharp initial selloff.

With sales up 47% year-over-year to a $14 billion run-rate and income surging 60% to $926 million, AWS has emerged as a crown jewel in Amazon's portfolio. It's hard to find fault with the unit's performance, particularly considering it has exploded from negligible revenue to such lofty heights in about five years. Still, analysts worry that continued price cuts will reduce growth rates this year, plus there are signs that cloud buyers are hedging their bets as offerings from Microsoft, Google, IBM and Oracle continue to improve (see my coverage here and here). (See chart at top.)

Cloud choices becoming more diversified

There are many ways of taking the pulse of enterprise cloud usage, but user surveys help get inside the macroscopic revenue and market share numbers.

One of the oldest cloud surveys comes from RightScale, the software vendor whose system helps centralize management of multiple cloud platforms. Like prior years, the 2017 edition polled people within its contact database, so it has some selection bias that likely skews to cloud-savvy organizations. Even so, only 20% were current RightScale customers, so we can take confidence that any skewing is likely minimal.

Within this universe, the survey found that 95%, use some form of cloud, with two-thirds having a mix of public and private services and another 22% solely in the public cloud. Despite the skew towards shared services, the number of workloads was almost evenly split, 41% to 38%, between public and private infrastructure, with SMBs showing a distinct preference for public cloud.


Where AWS may have cause to worry is in the cloud usage trends the RightScale survey unearths.

Although AWS is by far the most popular service, used by 57% of respondents with another 27% testing or planning to use AWS, its share has flatlined in the past year. Azure recorded a jump of 14 points to 34% and Google Cloud increased 5 points to 15%, with all others increasing in single digits. When measuring those already running, testing or planning to use a particular service, it's clear the market is a three-horse race with Amazon in the lead at 84%, Azure next at 67% and Google third at 45%. Among large enterprises with more than a thousand employees, the gap between AWS and Azure narrows to 11 points, 86-75%, when including the larger group of testers and planners.


Dominant incumbents are always vulnerable to new technologies that upset established customer behaviors, and in the cloud world, few things are potentially more disruptive than the use of containerized applications and infrastructure clusters.

Here too, AWS was an early adopter launching its Docker-compatible ECS service two years before containers became the hot IT topic of 2016.

Despite Google having arguably the most experience and technical depth with container infrastructure, its late introduction of GCE left Amazon to cultivate the market. The RightScale survey once again shows that being first pays off as 62% of its respondents are current or prospective users of ECS versus 39% for Azure's Container service and just 26% for GCE. Indeed, the results are even more lopsided when looking at current users, not just evaluators.


My take

AWS remains the preeminent cloud service and has done nothing to endanger its position, however if RightScale's survey is representative of a more cloud-savvy audience, the gap is closing.

Azure penetration is now 60% of AWS and Google Cloud adoption grew 50% in the last year. The tightening competition is the result of significant investments in new services, infrastructure and marketing by two companies with the technical resources to match Amazon feature-for-feature.

Microsoft benefits from its longstanding position as a core supplier to enterprise IT, while Google has profited from greater and more effective outreach to corporate customers under Diane Greene.

AWS still has a huge lead, although exactly how large is difficult to assess given the opacity of Alphabet's reporting as it relates to Google and the fact that Microsoft mixes Azure revenue with other cloud services.

As I've written before, both Google and Microsoft have kept up the pace of technology innovation such that in some areas, their services are arguably superior to the AWS equivalent. To counter, Amazon has wisely built its sales staff and partner alliances to win more enterprise business.

The combination of a virtuous flywheel that leads those trying core AWS features to adopt more and more services, along with the growing ecosystem of partners led one Wall Street firm to conclude it is impossible for Google to catch up.

I'm not convinced, but Google has much riding on its upcoming Next cloud event. In contrast, whereas Google is focused on cloud-native developers, Microsoft must convince enterprise customers that it can deliver on its hybrid strategy linking Azure public services with private Azure Stack infrastructure. Microsoft has a good shot at this as it has both the name recognition and partnering history upon which to draw.

Regardless, neither Google nor Microsoft have an easy task since AWS will continue to be an innovation machine with a widely recognized brand among the all important developer community. Both Amazon's major competitors have an opportunity to solidify themselves as strong alternatives in what is turning into a public cloud oligopoly.

A grey colored placeholder image