2 min read
Companies aren’t abandoning the cloud entirely, but many have scaled back their cloud initiatives. The main reasons, according to US market researcher IDC, are cost concerns, often disappointing performance, and inadequate security.
The Key Points at a Glance
- Companies are partially withdrawing from the public cloud, finding it often too expensive and slow.
- Common reasons include high costs, latency issues, security concerns, and compliance requirements.
- Over half of companies have already exceeded their 2024 cloud budgets.
- Only 8-9% plan to fully repatriate all workloads to their own data centers.
- Hybrid infrastructures combining on-premises, private, and public cloud are gaining significant traction.
When it comes to the public cloud, the initial enthusiasm among companies has somewhat waned. Many have begun a partial pullback because they find the public cloud too costly, underperforming, insecure, and difficult to manage. According to IDC, compliance issues add to these challenges.
However, the market research firm notes that this doesn’t mean companies want to abandon the public cloud entirely. Instead, many prefer a mix of on-premises, private cloud, and public cloud solutions, as heise online reports.
Only 8 to 9 percent of surveyed businesses plan to fully repatriate their workloads. This trend is particularly noticeable among large enterprises. The primary driver for the partial withdrawal from the public cloud is cost, with over half of companies already exceeding their 2024 cloud budgets.
AI and complexity are also driving up costs
Rising costs aren’t just due to higher provider prices – they’re also driven by the growing complexity of subscribed services, increased energy costs, and new technologies like generative AI. Yet both providers and enterprise users once promised that cloud adoption would reduce IT spending, as IDC points out. That assumption doesn’t always hold, as this case study from 37signals demonstrates. The company, known for Basecamp and HEY, saved $1 million in its first year after migrating from the public cloud to its own data center. It now expects annual savings of nearly $2 million – far exceeding its initial target of $7 million over five years. CTO David Heinemeier Hansson notes that the hardware costs for on-premises operations paid for themselves quickly.
When it comes to cloud performance, B2B customers surveyed by IDC cite latency issues as a major pain point, particularly for certain applications. For AI workloads, for example, an on-premises infrastructure often outperforms the public cloud.
Security is another key factor
After several incidents, many companies now question data security and data protection in the public cloud. The legal certainty of Microsoft 365 in German public-sector organizations, at least, remains a contentious issue. To err on the side of caution, many enterprises are adopting a multi-cloud strategy – yet this typically entails far greater administrative complexity. According to IDC, keeping control on-premises is often simpler for many companies.
Nevertheless, most enterprises continue to rely on the public cloud as a core component of their IT operations. The trend study conducted by IDC, however, clearly shows a shift toward a more hybrid approach. This model allows organizations to weigh critical factors – cost, performance, security, data protection, compliance, and manageability – on a workload-by-workload basis.
Frequently Asked Questions
Why are some companies pulling back from the public cloud?
For many, the public cloud has become too expensive, underperforms, raises security concerns, and proves difficult to manage. Compliance requirements are also a key factor.
What’s driving cloud costs through the roof?
Beyond rising provider prices, complexity, energy costs, and emerging technologies like generative AI are major contributors. Over half of companies exceeded their 2024 cloud budgets.
How much did 37signals save by moving away from the cloud?
The company saved $1 million in the first year alone and expects annual savings of nearly $2 million.
Why is on-premises infrastructure preferred for AI workloads?
According to IDC, on-premises infrastructure delivers better performance and lower latency for AI workloads compared to the public cloud.
What strategy are most companies adopting instead?
Many are turning to multi-cloud or hybrid strategies to gain better control over costs, performance, security, and compliance.
Editor’s Picks
- API-First: Why Modern Cloud Architectures Succeed or Fail at the API Design Stage
- Pretext: Can a JavaScript Library Solve the Browser’s 30-Year-Old Problem – or Is It Just Hype?
- AWS vs. Azure vs. Google Cloud 2026: An Honest Comparison for DACH Enterprises
More from the MBF Media Network
SecurityToday | MyBusinessFuture | Digital Chiefs
Source header image: iStock / sdecoret
