Cloud

How to Trim Costs by Reducing Cloudflation

Businesses are starting to get sticker shock on their cloud deployments. How can IT departments control costs?

Joel Keller spent more than a decade in IT before becoming a full-time journalist. He has written for The New York Times, Parade and others.

Like the cost of groceries, gas, rent and just about everything else, the cost of cloud computing has felt the influence of inflation and other market forces.

After over a decade of explosive growth in the use of off-premises cloud services and falling costs, especially from the big three — Amazon Web Services, Google Cloud and Microsoft Azure — cloud computing prices are on the rise. Because of this, businesses are finally feeling the financial pinch and looking for ways to lower their spending on cloud computing.

Tiny Increases Mean Big Bucks

Earlier this year, one major cloud platform announced price increases for some of its services. Effective Oct. 1, data requiring access at least once per month will increase in cost in some regions by 50 percent, from 1 cent per gigabyte per month to 1.5 cents; data requiring access about once per year will double in price for many customers, to as much as 20 cents per 10,000 operations.

While it may not seem like much, for enterprises that may have hundreds of terabytes of data stored in the cloud, those increases can be significant.

Click the banner below to receive exclusive cloud content when you register as an Insider.

Add that to the general increase in spending on cloud computing. For example, Synergy Research found that enterprise cloud computing infrastructure spending jumped from $100 billion to $130 billion between 2019 and 2020. Calculate it all, and it’s easy to see how a CIO might get sticker shock when looking over cloud infrastructure invoices.

Inflation Is Not the Only Factor

What is driving the rate increases? Inflation, including rising energy prices that affect everything from power to transportation, is just one factor. Supply chain problems are another: Taiwanese microchip manufacturers, which supply the vast majority of the parts for the devices that drive cloud infrastructure services, are still suffering from the supply chain issues that started when those plants shut down during the height of the pandemic.

Yes, the same semiconductor shortage that is making everything from cars to appliances increase in cost is also affecting the cloud infrastructure industry. Given the fact that Taiwanese companies account for a 66 percent share of worldwide chip manufacturing, those supply line delays have a big impact on hardware prices.

Also during the pandemic, demand for cloud infrastructure services sky-rocketed, with spending growing 19 percent in 2021 alone, according to Gartner, due to the demand brought on by remote work, online learning and other sudden changes to the way business was done — changes that aren’t going away.

DIVE DEEPER: Learn how cloud security posture management tools can help secure cloud services.

Businesses Need More Efficient Cloud Computing

Underneath the surface of the growth in cloud use over the past few years is a vast amount of waste. In its 2022 “State of the Cloud Report,” Flexera notes that organizations say they wasted 32 per-cent of the money they spent on cloud infrastructure.

Cloud waste essentially comes from an overestimation of what cloud resources a particular project needs. Cloud waste includes problems like planning too long of a time frame for a particular cloud-centric project, over-provisioning resources, and a general lack of maintenance and optimization, especially for data and projects that no longer need to be in active storage or that utilize a lot of CPU capacity.

How to Reduce Cloud Costs

There are some fundamental things you can look at to reduce cloud costs:

Optimize cloud purchasing between groups. Expedia’s vice president of development and runtime platforms found that each company in its portfolio had purchased its own cloud-based development platform. By standardizing on a single platform during the pandemic, the company experienced a huge cost reduction.
Choose the right kind of resources for your needs. The major cloud platforms offer “reserved” or “spot” options, which allow businesses to purchase a certain amount of computing instances at discounts of up to 90 percent compared with on-demand instances. For workloads that can be interrupted and require server instances for a particular period of time, and if you can work with the configuration standards of those instances, there can be big savings.

Find and remove unused resources. If a temporary server instance was created for a project and never removed, or if cloud storage was used for a project but never deleted as the project wound down, now is the time to jettison them.
Use cloud providers’ tools to prevent overprovisioning. In the data center days, server administrators often planned for a lot of hardware headroom to meet peak demands and anticipate growth. But with tools such as autoscaling, load balancing and more, cloud providers can help organizations respond to demands while preventing low utilization of server instances and storage.
Right-size existing project needs. The major cloud providers all offer right-sizing tools to help IT managers decide whether to switch cloud configurations to optimize usage and utilization. Because of all the configurations offered, right-sizing isn’t for the faint of heart. But increasing utilization while reducing costs is a major motivating factor to try.

The cloud’s benefits are undeniable, but being cheaper than on-premises technology isn’t necessarily one of them. As businesses transition more workloads to the cloud, and as cloud computing costs rise, it’s vital that technology leaders keep a watchful eye on their total spending and make adjustments when necessary.

Illustration by LJ Davids