There has been a lot of discussion, including Part I of this series, about AWS’s relentless leverage of scale to drive down pricing in cloud services. What is much less talked about is how its consumer platform drives its cloud leadership.
The reason why AWS will continue to lead the cloud services market – and Iaas in particular – is the leverage it gains from its massive cloud consumer business. Its extensive mesh network serves as an infrastructure kitchen in which AWS can test and consume its own resources on an unparalleled scale.
The scale of this infrastructure is only approached by two of its nearest contenders in cloud services, Google and Microsoft. Not coincidentally, both also have enormous cloud-based consumer driven businesses. In Google’s case, search advertising, and in Microsoft’s, it’s games.
By eating their own cooking these cloud providers can continue to drive down the cost of provisioning and deploying infrastructure. As AWS is demonstrating, this provides enormous pricing leverage. While some may view this as an attempt to marginalize all competitors early in the evolution of cloud services – which to an extent it is – others may view it as a company in relentless pursuit of reducing the cost of computing to all of its customers.
This is a competitive advantage that no other cloud services provider can match, whether through a proprietary model or an open source one. Owing to its retail service “mentality”, few cloud vendors can maintain a mission of reducing customer cost while continuing to build more flexibility into the architecture.
Why major players choose AWS
It is for this reason why Netflix runs on AWS. Companies that are already running their business on AWS, as well as larger enterprises that are considering moving services to the platform (i.e. GE) may want to take notice: a tool used by the largest AWS customer for visibility and operational support may be quite valuable. Tying in to our earlier post, Ice uses data supplied by AWS APIs to gain granularity into usage patterns.
AWS’s competitive advantage is also why two years ago the CIA chose AWS’s “superior technical solution” over IBM for a $600 million private cloud contract. IBM’s first response was to challenge the decision using its leverage at the GAO. Its second response was to acquire SoftLayer.
While a clever tactic, SoftLayer will not get IBM any closer to competing with AWS – or Google and Microsoft for that matter – in cloud services. Why? Because IBM still lacks a massive cloud consumer platform. It will, however, help IBM sell more software and Global Services contracts.
And GE’s decision to run its Industrial Internet initiative on AWS should be a clarion call to organizations everywhere. It should also serve as a wake-up call to all cloud service providers, including the open source consortiums. We suggested in another post that big data analytics was going to be GE’s new Six Sigma. Through new partnerships with AWS and Pivotal, GE is jumping into the pool with a big splash.
Open source community take heed
Only open source can provide the scale to compete with AWS, even without a direct consumer-driven platform. However, as OpenStack continues to evolve its ecosystem of tools vendors, the lock-in risk we addressed in Part II of this series becomes an inhibiting, rather than an enabling factor. It also makes the decision to go with AWS that much easier to justify.
Netflix’s decision to open source tools like Ice, Isthmus for managing elastic load balancing across AWS regions, and Zuul, to manage the interface between its own and AWS’s APIs makes a lot of sense. Providing an open source platform that makes life easier for DevOps teams is a first step toward wider deployment of these tools across different cloud services platforms.
