Modern Stack

Insight on building and supporting cloud apps

Minerva Studio - Fotolia

The enterprise will kill cloud innovation -- but that's OK

Hyperscale cloud providers grew incredibly fast and created a whole new way of doing business. That's great, but it's also a problem that needs fixing.

If you're not fully invested in the cloud today, you're guilty of IT malpractice. Of course, I'm exaggerating. But many IT pros feel this pressure from business leaders at a time when everyone seems to be talking about cloud innovation.

At Google's 2018 Cloud Next conference in San Francisco, the company proudly proclaimed that its cloud was "enterprise-ready." This pitch clearly aimed to ease security and stability concerns and attract large accounts at a time of increased competition among the big three cloud providers. It also underscored a disconnect between cloud providers' prior marketing messages and the reality that enterprise IT pros face.

Just weeks before this conference, Google's cloud business faced fierce criticism after the provider's automated fraud-prevention systems pulled the plug on a customer's business-critical project without warning. The customer's published account -- bluntly titled, "Why you should not use Google Cloud" -- prompted a public apology from Google and promises to change. On one level, this was a cautionary tale of automation run amok -- a warning that even advanced AI can't replace a human.

It also pointed to a broader problem facing the world's largest cloud providers. Cloud innovation moves too fast for many large customers.

Google's and Amazon's cloud services are among the most innovative and disruptive technologies to hit the IT industry in years. Yet neither company was born as an enterprise IT technology provider. In some ways, this accelerated the services' early growth. The companies' early success wasn't tied to rushing out incomplete products to generate revenue; they could fail fast with minimal consequences, and they certainly didn't have to maintain legacy systems. Pay-as-you-go instances accessible to anyone with a credit card fueled growth and led to a sort of commoditization of compute resources.

This accessibility, however, created a unique problem for cloud providers. Cloud instances that began as pet projects, developer experiments or shadow IT have quickly graduated to critical applications. The key for providers is not just to offer business-level support at an additional cost, but to rethink all their services as business-focused. There's only one way to accomplish this goal: slow down and spend more.

Hyperscale providers turn a profit by maximizing resource use and relying on automation to minimize human costs. As the Google example shows, large cloud providers must learn to shift gears if they want to retain business customers. Corporate accounts need more support, customization and stability -- as opposed to agility -- than the typical cloud provider offers individual consumers. Providers will need to continue to evolve to target large businesses, shifting the way IT pros think about the cloud.

This shift is happening now. In 2017, Amazon began working with VMware to offer up entire hosts running VMware's ESXi hypervisor -- a service that feels like traveling back in time, compared to the agile, flexible, pay-per-second model AWS popularized. It has undoubtedly been a critical experience in learning how to cater to enterprise customers.

The key for providers is not just to offer business-level support at an additional cost, but to rethink all their services as business-focused.

On the surface, it may seem odd that the leading public cloud provider, known for pioneering new services, should team up with an established company peddling yesterday's software. But it's not just about the technology. VMware, and many other established IT vendors, have the experience grooming enterprise customers that cloud providers lack.

In the coming year, I expect cloud providers will deepen relationships with a range of legacy vendors. We'll also see cloud innovation slow and traditional cloud services look much less exciting. IaaS will evolve into a stable platform to better suit legacy workloads, with fewer updates that could disrupt customers. Within 10 years, the monopoly today's hyperscale providers have on IT innovation will fade as they find themselves investing more to maintain the status quo and preserve long-term relationships. It's the natural cycle of IT technology, and it'll be to the benefit of customers who want the flexibility of the cloud without the risk and disruption.

Before we know it, some other technology will emerge as the next big thing, moving fast, breaking things and exciting IT pros the same way cloud innovation did -- just remember that risk always accompanies innovation.

Article 4 of 6

Dig Deeper on Cloud deployment and architecture

Data Center