VP of Cloud Services at Unisys.
The pandemic has hit sectors such as higher education (paywall), retail, transportation and travel especially hard. But no industry has been immune. That has organizations of all stripes doing everything they can to survive this challenging time and emerge stronger from it.
As the vice president of cloud services at a company that offers cloud acceleration and optimization solutions, I’ve found that one way they can do that is by optimizing their cloud deployments to cut down on cloud sprawl and wasteful spending. Cloud optimization may not be the first thing C-level leaders are thinking about right now. But it can help eliminate waste at a time in which many business leaders are seeing their budgets cut. And it can free up funds that organizations can then reallocate to innovate and better position them for the future.
We’re not talking about pocket change here. Research compiled in B2C by an employee of a cloud spend management platform indicates that organizations will throw away more than $17.6 billion of their cloud spend this year. Of that, $11 billion will be wasted on idle cloud resources. The other $6.6 billion will be squandered on oversized resources.
You may not notice your wasted cloud spend.
These numbers highlight the value of keeping a watchful eye on cloud spending. Enterprises need to keep track of and constantly evaluate cloud choices just as they do with other budgets and resources.
Think about all the data center controls people have developed over the years. There is so much accountability about floor space use, what hardware you have, how it’s depreciating and how to plan capacity for future demand.
But it’s easy to lose track of your cloud deployment. Enterprises have theoretically unlimited capacity on demand with cloud regions across the globe. And when shadow IT is involved, wasteful cloud spending could be hard to spot. You get these huge bills, which you likely pay without realizing dollars are going down the drain.
Wasted cloud spending comes in many forms.
Using public cloud resources opens up the opportunity for a variety of wasted cloud spending, including overprovisioned compute resources. Calculating the cost of a virtual machine (VM) is a surprisingly complex exercise based on the class of the virtual machine, in which cloud region you’re hosting it, et cetera. But large VMs (those with more CPU cores and memory) are generally more costly than smaller ones. If you have VMs that are far larger than what you need, you are throwing away your budget. Yet as one TechTarget article explains, admins seldom, if ever, right-size VM resources — even if there is an overage. It’s kind of like getting the wrong-sized house and spending extra for the upkeep.
You might also have zombies, including unattached and unused storage and VMs that your company isn’t using. And you’re just racking up extra costs and resources that you simply do not need.
Targeted FinOps and cloud management bring order to the chaos.
You should consider business outcomes when you’re trimming your cloud spend. Let’s consider a few examples. Maybe you transfer a big load of data on a nightly basis. But do you really need to transfer all that data every night? Or can you trim the amount of data you’re sending to reduce cloud spending on data transfers and data storage and really drive your intended business outcomes?
Or perhaps you work for a business that has multiple subsidiaries and has grown through acquisition, and you have created a data lake to pull the business metrics together. Over the years, your data lake has grown. At some point, someone needs to decide when to do archiving. How do you decide when to do that and what data to archive?
The total cost of ownership reduces as you refactor legacy applications and adopt more cloud-native services. Are the right investments in application modernization being done to achieve this?
If you have a FinOps practice that brings together cloud operations and financial expertise, you will have the governance to ask and address such questions. You should understand how critical the application is and the business outcome it delivers. The FinOps office can make buying decisions that can dramatically alter the cloud bill — e.g., as a Unisys colleague explained, they can buy reserved instances that can lead to discounts. If you make a minimum spend commitment to qualify for an enterprise discount program offered by the cloud provider, you can achieve significant savings. Most organizations I’ve worked with have FinOps efforts, but many of them are not yet focused on the cloud.
You may also want to adopt a cloud management platform that has quotas and approvals for provisioning (something my company offers, but others do as well). Cloud management platforms and FinOps can help improve your cybersecurity posture and lower costs by spotting spikes and validating demand peaking. That reduces the potential for bad actors to use your cloud resources and cloud dollars for purposes like mining cryptocurrency, which is what happened to Tesla.
Cloud optimization has peak potential when you incorporate artificial intelligence and automation.
Many cost optimization tools look at current capacity and utilization. But they are not always smart enough to adjust to factors like seasonality and dynamic cloud environments. They may also lack the ability to predict cloud requirements so organizations can acquire precisely the right resources.
By incorporating AI into your cloud optimization, you can right-size your cloud resources, know if you have the right storage and other cloud-based assets attached and predict your future cloud needs.
When you pair AI with automation, you can spin up and take down cloud resources as your workloads demand. With such hyper-automation, you can take optimization to the next level. You can cut costs without eliminating projects. You may even be able to increase the number of projects.
Using your money wisely is always a good idea, but it’s especially important as we all work to survive and thrive in this new era. By leveraging intelligence and automation to optimize your cloud spending, you can free resources to make your business stronger and more resilient.