You are here

You are here

Fix these problems or they'll drive up your cloud service costs

public://pictures/seandaken.jpeg
Sean Daken Director, Research, Innovation & Emerging Technology, Trace3
Navigation tools
 

The last decade saw a transformational move to public cloud infrastructures for most large enterprises, and today many companies run their critical business processes and applications on a mix of cloud infrastructures and supporting services.

In fact, 94% of large enterprises now use cloud services, according to Flexera's "RightScale 2019 State of the Cloud" report. But CIOs are now taking a more measured approach to cloud cost containment and optimization, as wholesale moves to the cloud have proved to be more costly than expected.

Enterprise IT leaders need to take a strategic view of their cloud environments and service footprints to better determine how to manage their cloud spending, now and for the future. 

Why it's difficult to forecast cloud costs

Enterprise IT budgets have swelled in recent years. The most recent expected enterprise IT budget increase is 18% year over year, according to a Spiceworks survey, "The 2020 State of IT: The Annual Report on IT Budgets and Tech Trends." 

There is a perception among IT executives that cloud costs are outpacing their budget allocations and ultimately spiraling out of control. Many organizations that have moved applications and workloads to the cloud have done so with haste, overlooking the finer points of cloud cost provisioning.

The frenzied push to convert CAPEX into OPEX results in poorly forecast costs, and, without adequate cost containment strategies, cloud services costs often spin out of control.

In general, organizations tend to neglect the soft costs of doing business, and this is also true when moving to the cloud. Enterprise IT teams regularly discount the length of time and the degree of change management required for a successful cloud migration.

IT teams often lack an adequate understanding of the downstream implications of a cloud migration, including but not limited to the skills required to be successful, new risks, and additional compliance reporting requirements.

The move to multicloud and other complexities

Many enterprises are moving from single- to multicloud environments. Some 84% of enterprises are using multiple public clouds, and enterprises with a hybrid strategy are above 50% and growing, according to the "State of the Cloud" survey.  

Unfortunately, only about 50% of enterprises are using cloud management platforms, according to 451 Research. And enterprises without a cloud management platform can expect poor visibility into their cloud environments, which can lead to runaway workload costs.

Another issue around cloud deployment is communication problems among different teams. In an article titled "DevOps dictates new approach to cloud development," David Linthicum of Deloitte Consulting described the situation well: "What's most interesting is that the cloud isn't really driving DevOps; rather, DevOps is driving the interest and the growth of cloud." The conclusion is "DevOps is wagging the cloud computing dog, not the other way around."

For example, if a quality assurance (QA) team provisions cloud resources for internal testing, but those resources remain active when not being used, that's a problem. Without effective communication, the DevOps team may deprovision those cloud resources when the QA team needs them, leading to frustration and bottlenecks.

Where do we go from here?

Although it's true that IT execs perceive that cloud costs are spiraling out of control, the reason the situation has arisen isn't because the underlying service costs are increasing. The absolute costs of computation, information transfer, and data storage (per unit of measurement) fell exponentially for nearly all of the 20th century; this situation has continued into the present.

This is thanks to the economic phenomenon of experience curves captured in Moore's Law and the lesser-known but related laws of Gilder and MetcalfEdholm, and Kryder.

So, contrary to popular belief, cloud costs are not rising in absolute terms. Don't fall for that. Instead, CIOs should be asking the following questions:

  • Which factors are driving up my costs?
  • How can we navigate the cost optimization journey?

Cloud cost optimization best practices

Enterprises should take these three steps to more effectively manage their cloud services.

Move workloads and applications to the cloud on a piecemeal basis

Consider doubling down on pilot projects to better ascertain the true costs of cloud services. When done right (small-scale, objective measures of success), pilots and a proof of concept (PoC) can reduce downstream costs by allowing your organization to traverse the learning curve with lower risk, turning those "unknown unknowns" into "known unknowns."

Even if your enterprise has already gone down the path of cloud adoption, doing pilot projects and PoCs can prevent additional unexpected cost proliferation.

Implement cost optimization and governance tools 

Ensuring visibility into the cloud environment is a principal lever in reining in costs. Moreover, establishing governance controls is critical to ensuring that cloud infrastructure policies are defined and enforced.

Many cost optimization tools have some built-in governance capabilities. Unfortunately, there’s no silver bullet—no single tool currently covers everything well. The recommended approach is to select a cost optimization tool in combination with a governance and security tool. This is also where policy automation comes into play

The related challenge is deciding which group within an enterprise should select and implement these tools. This is typically a tug of war between finance and IT. There is an inherent tension between those teams in most organizations, and they have difficulty finding common ground. Which path an organization takes depends on the use cases being addressed and how the organization is structured.

Improve communications

Finally, develop rules and communication processes with your DevOps team. DevOps teams inadvertently drive cloud services costs, so having effective cross-team operating and communication mechanisms is critical to ensure the right work is getting done without breaking the bank.

Having a cloud strategy is key to success

Containerization and serverless computing will be significant drivers of public cloud growth in the next few years. To take advantage of these emerging trends in cost-effective ways, IT leaders must tackle the basics of cloud cost optimization and governance now. These issues are increasingly becoming as important as addressing the cybersecurity challenge faced by enterprises.

To adequately address the challenge of cloud cost optimization, your organization may need to adopt multiple, complementary tools and governance processes, which is why first defining your cloud strategy is key.

Given the number of vendors in cloud cost optimization and governance, IT leaders should focus on mapping out a cloud strategy for specific use cases. This includes identifying solutions that address your organization's specific use cases and then selecting two to three vendors for a deeper comparative dive and PoC.

In the absence of an effective cloud strategy, cloud costs will continue to rise. It's up to enterprises, their CIOs, and their strategic technology partners to rein in those costs, and to do so in a way that doesn't stymie innovation or digital transformation initiatives.

Keep learning