VentureBeat presents: AI Unleashed - An exclusive executive event for enterprise data leaders. Network and learn with industry peers. Learn More
Over the past decade, cloud adoption has become the rule, not the exception. And yet, many companies that have embraced the cloud are feeling the acute burden of a spike in spending. In other words, cloud usage may be costing many businesses more than they are actually saving.
Despite that, 69% of businesses globally have accelerated their migration to cloud computing over the last six months and end-user spending on public cloud services worldwide is predicted to swell to nearly $482 billion by the end of this year. This exploding cloud-spend is often based on the assumption that a company’s extensive investment in the cloud today will ultimately make computing a more economical endeavor in the future.
The market is down, companies are looking to cut costs, and many are reassessing their cloud spending. But cloud reliance is so vital to business continuity today that many organizations are reluctant to touch their often-bloated cloud budgets, even if they know there is fat to trim. This includes startups with the huge expense of overly leveraged cloud usage but lack of leadership insight into where to make cuts.
Here are some tips for enterprises to get a better sense of their cloud costs and ultimately mitigate the wasted spend.
An exclusive invite-only evening of insights and networking, designed for senior enterprise executives overseeing data stacks and strategies.
Understand basic cloud costs — and waste
The majority of cloud costs can be broken down into three primary components — service, utilization and unit price — all of which can yield waste if mismanaged.
Service refers to the cloud service provider an organization uses for any given function (that is, AWS, Microsoft Azure, Google Cloud). It also encompasses the overall technical specs of the service including instance family, instance size, processor, and operating system. “Compute” and “storage” are two service types that are responsible for major costs. It is challenging to alter service decisions once cloud usage is underway, so organizations must have a keen understanding of this component from the get-go.
Utilization considers how well a given organization actually uses and implements the service chosen. This includes the volume of the service being used, how much data is regularly being transferred across cloud platforms, and the effectiveness of “rightsizing.” Unnecessary costs can arise when instances are not maximized for efficiency — you pay for the full instance even when you don’t use it fully — as well as when unused “zombie infrastructure” is left unchecked.
Unit price refers to the amount you pay for service usage. Major service providers such as AWS have a standard on-demand unit price but also offer lifelines like savings plans, committed use discounts, and enterprise discounts. All of these reservations can be negotiated with cloud service providers and stand to reduce cloud-spend overhead.
Manage cloud cost
At its core, cloud cost management boils down to a three-pronged methodology of visualization, optimization and monitoring.
Oftentimes, organizations will embrace cloud capabilities simply for the sake of “digital transformation,” adopting this transformative tech without a sufficient understanding of every dollar spent or what function it will serve. But this lack of visibility into cloud utilization and costs quickly becomes a major problem. This is particularly the case for businesses that use multiple cloud services or platforms across the organization, as each respective cloud vendor only provides visibility within their own offering and getting a holistic picture of cloud spend across vendors becomes extremely challenging.
Organizations should start by visualizing a “map” of the following conditions: which people or teams are using which respective services; how much computational volume is being utilized by each respective team or employee; and the bottom-line breakdown of how unit price is impacting total cloud budget cost.
This process of visualization must transcend internal siloes and touch on every facet of the organization where the cloud is utilized. If each team is using a different dashboard, each its own specialized “language,” cloud processes will become like the Tower of Babel — no common tongue to be found. Only a birds-eye assessment of all platforms and dashboards at once will allow organizations to communicate properly in order to identify where they are overspending on the cloud.
Once the process of visualization is complete, IT teams must try to optimize cloud functions along the following lines: improve instance-to-workload alignment; size down overprovisioned infrastructure and terminate zombies — all while making the most of any available reservations or discounts.
On top of visualization and optimization, there must also be an ongoing process of monitoring for possible configurations that might improve instance-to-workload alignment, for anomalous utilization intervals and for impacts on bottom line costs. Even once cloud costs are under control, enterprises still must be vigilant to ensure that costs don’t creep back up. You can always spot further opportunities for budget optimization.
Don’t let the cloud rain on your success
There’s a good reason for the growing ubiquity of cloud usage: With proper implementation, the cloud can make work easier and more collaborative, improve efficiency and communication across an organization, guarantee flexibility, scalability and business continuity — and, at its best, save money.
But the digital cloud is not unlike the ones in our sky — dynamic, intangible, ever-shifting, easy to lose sight of… look away for a moment and it may have taken on an entirely new form.
Cloud-reliant companies — eventually, all companies — must assess services, utilization, and unit prices across the enterprise through a process of visualization, optimization, and ongoing monitoring.
That way, they can guarantee bright skies even in tough times.
David Drai is CEO and cofounder of Anodot
Welcome to the VentureBeat community!
DataDecisionMakers is where experts, including the technical people doing data work, can share data-related insights and innovation.
If you want to read about cutting-edge ideas and up-to-date information, best practices, and the future of data and data tech, join us at DataDecisionMakers.
You might even consider contributing an article of your own!