For businesses pursuing innovative ways to use data analytics, the cloud seems like an ideal option for data management.
Cloud spend, however, can easily obliterate budgets without guardrails and oversight in place.
With the C-suite closely watching technology budgets, optimizing management practices and changing cloud operating models will help data and technology leaders reign in spend, according to Adam Ronthal, VP analyst at Gartner, during the Gartner Data and Analytics Summit 2021 on Wednesday.
Cloud is the way to go for future data management — especially for organizations anticipating the addition of edge computing and IoT technologies — but almost every attribute that makes cloud attractive to the business contributes to overruns, according to Ronthal.
Rapid provisioning and agile deployment capabilities cut down on time to procurement and increase efficiencies. But these strategies also mean variable usage costs, which leads to less predictable budgets, according to Ronthal.
Cloud decision-makers estimate about one-third of their organization's cloud spend is wasted, according to Flexera's 2021 State of the Cloud report. Unused discounts and lack of contract customization often fuel the overrun.
Plus, with so many options for getting to the cloud, "it's not always inherently obvious which one of these is going to give me the best cloud efficiency and the best cost efficiency, so cloud also represents a fundamental increased complexity," Ronthal said.
In 2020, 42.6% of organizations overran their cloud budgets, according to a Pepperdata report. Increased complexity often leads to increased costs to manage metadata, governance and integration.
To solve the cloud cost problem, organizations can rethink their existing structures for data management in the cloud. Ronthal recommends taking advantage of cloud's unique cost-saving provisions, consistent oversight of the cloud environment and a FinOps approach to cloud cost.
Optimizing cloud policies and operating models
Effectively managing cloud costs relies on changing an organization's approach to data management policies and operating models.
The lift and shift model of moving to the cloud undermines long-term opportunities for cost benefits, according to Ronthal.
"If I simply lift and shift and do everything the same way that I was doing in the on-premises world, I'm not going to be able to take full advantage of the cloud," Ronthal said. Over a four to seven year period, the costs will be roughly equivalent.
Instead, businesses can take advantage of cloud's unique cost-saving offers. Using cloud-native capabilities, provisioning compute based on use and dynamic elasticity can all help organizations avoid the cost overrun, according to Ronthal.
"If we can make our workloads use fewer resources, consume resources more efficiently, we actually wind up spending less," Ronthal said. "We have to engage with a fundamental design philosophy shift in the cloud."
This isn't a static shift. Cloud consumers engage with the vendor and the price performance over time so that as soon as there's an indication that cost overrun could occur, it can be corrected, according to Ronthal.
Pricing models can also be aligned to business needs, according to Ronthal. Depending on the cloud, it could be based on resource use, consumption or other models that leadership can tailor to the workload.
Ronthal recommends a FinOps model to cloud deployments that manages the financial operations by assessing workloads based on cost and value. FinOps helps organizations "keep in mind a future vision" that iterates for long-term cost optimization.