Dive Brief:
- IT teams anticipate a shift in enterprise cloud strategy over the next two years, according to a Rackspace Technology survey of 1,400 IT decision-makers published Tuesday. More than 9 in 10 respondents said they expect significant changes.
- Nearly half of enterprises planning a strategic realignment aim to double down on hybrid capabilities to advance IT respondents, the survey found. In addition, 1 in 5 intend to invest in private cloud solutions.
- AI is fueling the strategy shifts, Rackspace said. “As AI continues to drive rapid advancements, IT leaders are reevaluating their approach to workload management,” said Srini Koushik, president of AI, Technology and Sustainability at Rackspace. “Instead of updating existing infrastructure, they are designing entirely new cloud strategies to meet evolving demands.”
Dive Insight:
IT teams reassessed data architectures, reviewed security risks and realigned skills around AI as enterprises moved from piloting use cases to adopting copilot assistants, coding tools and agentic solutions. The infrastructure supporting these technologies came under scrutiny, too.
Public cloud delivered large language models to the enterprise and gave developers ways to build generative AI tools. Despite the benefits, data privacy concerns and the potential cost of running compute-hungry applications on hyperscaler platforms triggered a rethink.
More than two-thirds of respondents to the Rackspace survey said their organization has considered repatriating workloads from public cloud to on-prem infrastructure. Half of respondents cited security and compliance as the primary concern while 44% pointed to the cost implications of running generative AI in public cloud.
Many organizations experienced sticker shock as AI proliferated in the cloud and bills rose accordingly last year, according to a survey commissioned by technology expense management solution provider Tangoe.
“The costs in public cloud rose so much that people are saying we really need to evaluate where things go,” Chris Ortbals, chief product officer at Tangoe, said. “There's still a huge need for workloads that aren't fit-for-purpose in public cloud for any number of reasons.”
Fully modernized applications and workloads that have bursts of activity punctuated by periods of low usage are typically optimized for public cloud’s billing structure. Big data can run on public or private cloud, depending on an organization’s needs.
Nearly 9 in 10 respondents to the Rackspace survey said they transfer workloads between public and private cloud seamlessly. As hybrid strategies incorporate more than one cloud provider, IT teams are also finding it easier to move data between multiple hyperscaler deployments, too, according to Rackspace.
AWS, Microsoft and Google Cloud removed some egress fees last year, making data transfers less costly. Oracle made strides toward multicloud interoperability, as well, signing deals with all three of its largest hyperscaler peers over the course of the year to put Oracle database hardware in their data centers.
