Ever since Basecamp announced late last year that they were (mostly) pulling out of the cloud, there’s been a spirited debate about the failure of cloud computing to deliver on hyped-up promises of miraculous cost savings. Like a low-calorie cheesecake, cloud computing was supposed to give us everything we desired with minimal effort, all while trimming our IT infrastructure expenses. Or so some claimed, and so some believed.
The truth? Cost savings is (and always has been) the worst reason to migrate to the cloud.
While there is a small subset of have-your-cheesecake-and-lose-weight-too scenarios in cloud computing, investing in the cloud for cost savings will often disappoint. It’s not how the cloud should ever have been sold, and chasing marginal cost reductions has pulled focus away from the cloud’s exponential potential.
Cloud instances are not magic boxes from which we can conjure oversized profit margins with infrastructure incantations. But, implemented wisely, cloud computing can still transform your business.
I work with a lot of CIOs who are:
Considering a migration to the cloud or trying to convince their CEOs to support the same.
Attempting to extract more cost savings from their completed migration in order to justify the incurred expenses.
Realizing that cost savings are not the primary benefit of the cloud and are trying to figure out how to leverage the greater potential of cloud-based deployments.
These CIOs come to me and my company with some fair, wise, and important questions. Here’s how I answer them.
Will investments in cloud computing reduce my IT infrastructure costs?
You can guess my bottom-line answer here: Cloud computing may not reduce your infrastructure costs, or, if it does, the savings may be marginal and not enough on their own to justify the investment.
That said, there are definitely more and less cost-conscious ways of implementing cloud infrastructure, and it’s not unusual for my company to discover that a new client has previously implemented their cloud in the most expensive way possible. There are best practices for optimizing cloud infrastructure spends so that expenses don’t get in the way of the greater value cloud computing offers. I’ll say more about those practices below.
David Linthicum, the author of An Insider’s Guide to Cloud Computing, recently wrote an article for InfoWorld that addresses this question. In “Making a new business case for cloud computing,” he writes, “The most significant value of cloud computing is rarely found in cost savings, although they sometimes do occur; it’s about delivering the more critical business values of agility and speed to innovation.”
Agility and speed to innovation.
Cost savings, he suggests, is a “hard value” that’s easy to define and measure but not reliably provided by cloud computing. Agility and innovation are “soft values” that are more difficult to define and measure but ultimately more valuable and likely to be delivered.
Will your organization’s success be served by greater agility and speed to innovation? It’s not a rhetorical question; not every company is in the business of adapting quickly and breaking new ground. But if yours is, the truly game-changing benefit of cloud computing comes from jettisoning the ponderous pace and long-term capital commitments of on-prem computing in favor of resources that can be provisioned in seconds and abandoned at will.
Cloud infrastructure can adapt to your company’s changing needs and new ideas at the speed of your best ideas.
OK, so cloud computing is not necessarily a hard cost savings, but how can I implement it cost-effectively?
“Lift and shift” ain’t it. If you move from bare metal to dedicated VMs as though nothing fundamental has changed, you’re probably not going to see cost savings. You may spend even more.
Cloud computing is a paradigm shift that requires us to think differently about system architecture; you have to stop treating your cloud like a data center. It also requires an operational shift in how we all work together to craft high availability systems and applications while controlling costs.
Don’t set up one-to-one applications deployed each to a dedicated VM. Instead, treat your cloud as a platform on which infrastructure is deployed via Terraform (or another infrastructure as code framework) to meet your dynamic needs. Containerize everything and use clustering platforms like Amazon’s ECS for orchestration. Utilize these and other cloud-native tools to make the most of the cloud’s ability to provide the services your company needs when you need them without all the carrying costs of boxes you don’t presently require.
Contrary to another popular myth, architecting cloud computing infrastructure isn’t any simpler than designing a data center. It requires just as much (maybe more) careful thought and planning to implement effectively. And while it solves some of the hazards, limitations, and expenses of on-prem computing, cloud computing replaces these with its own novel challenges and risks.
Rising to these challenges is worth the work, however, for that “agility and speed to innovation” you’ll unleash.
In practical terms, this means equipping your application engineers to develop their own infrastructure. With the ability to implement and maintain infrastructure as code, developers no longer need to rely or wait on a DevOps engineer or team to build out what they need right now.
In fact, if you architect your cloud infrastructure well, you no longer need a dedicated DevOps team. Transition those engineers instead to a platform engineering team to pave the road to production, then empower your application engineers to innovate on that platform at ludicrous speed.
Cloud computing infrastructure isn’t automatically cheaper, simpler, or superior, but, if you do it right, it can power your company’s revolutionary potential.
CIO, Cloud Computing