For over a decade, cloud-first was the dominant enterprise IT strategy. Organizations migrated applications, workloads, and data to public cloud platforms, drawn by promises of cost savings, scalability, and flexibility. The philosophy was straightforward: offload infrastructure management and focus on core business functions.
The reality proved more complicated.
Cloud bills that started as predictable monthly expenses evolved into complex, multi-line items. Data egress fees, API charges, and unpredictable compute costs added up fast. Moving sensitive data across jurisdictions introduced compliance exposure under GDPR, HIPAA, DORA, and data sovereignty regulations. And for many organizations, the sheer volume and variety of unstructured data — emails, documents, file shares, collaboration logs — created risks they had underestimated at the outset.
The Repatriation Trend
2026 marks a strategic inflection point. According to a Barclays CIO survey, 86% of enterprise IT leaders plan to move at least one workload off the public cloud, the highest rate ever recorded. IDC data puts annual repatriation activity at 70% to 80% of companies. A separate survey found that 97% of mid-market organizations plan to shift select workloads away from public cloud environments.
There are three forces driving this shift:
- Cost control. Approximately 27% of cloud infrastructure spending goes toward underused resources, according to Flexera’s State of the Cloud Report. Steady-state workloads do not benefit from pay-as-you-go pricing. 37signals documented roughly $1.3 to $1.5 million in annual savings after repatriating.
- Regulatory pressure. DORA is now live across the EU. In the U.S., HIPAA, GLBA, and FERPA require provable control. Regulators want audit logs and documented evidence of data governance.
- AI workload economics. Organizations running consistent AI workloads are finding that dedicated on-premises infrastructure often proves more cost-effective than paying premium public cloud rates for continuous, predictable GPU compute.
Hybrid architecture is emerging as the strategic sweet spot. Pairing cloud elasticity with on-premises control and predictability gives organizations the flexibility to place workloads where they perform best. With 92% of IT leaders expressing confidence in on-premises cybersecurity versus 78% in fully cloud-based environments, security adds further weight to the case for hybrid.
As AI capability accelerates, proprietary data has become a primary competitive differentiator. Organizations that store sensitive data in public cloud environments face a compounding risk: not just regulatory exposure, but the potential loss of the strategic asset that will define competitive advantage in the years ahead.
What Gets Overlooked: The Data Itself
Most enterprise data environments carry significant volumes of redundant, obsolete, and trivial (ROT) files. These accumulate across years of storage activity, migration cycles, and collaboration tool sprawl. ROT generates cost, compliance exposure, and risk at every stage of a cloud migration.
Governing and classifying data before migration delivers compounding benefits:
- Reduces migration scope and one-time transfer costs
- Lowers ongoing cloud storage fees, which run higher per GB than on-premises environments
- Minimizes compliance exposure by removing sensitive data that has no business moving to the cloud
- Ensures AI systems ingest clean, classified data rather than unclassified noise
Sensitivity classification is the critical enabler. Organizations need visibility into which data is high-value, which carries regulatory obligation, and which is accumulated ROT. Without that foundation, any migration carries the same risks forward into the new environment.
Migrate the Essence, Not the Sprawl
In-place data management offers a fundamentally different approach to hybrid cloud strategy. Rather than executing a blanket lift-and-shift migration, in-place management extracts and indexes the “essence” of every document — metadata and content — without copying or moving the original file.
With this method, organizations can migrate only the essential intelligence while the original documents stay where they are. High-value documents like contracts can still be selectively archived. From a unified platform, organizations govern across all repositories simultaneously:
- Records management: classification, retention, and defensible deletion across all data sources
- Policy enforcement: automated tagging aligned with regulatory requirements, with the ability to update policies without re-indexing
- Risk management: remediation of sensitive information and elimination of ROT files across on-premises and cloud sources
- AI readiness: curated, governed data delivered to AI pipelines without relocating massive data volumes
This model dramatically reduces storage burden and makes hybrid deployments operationally viable at the enterprise scale. Organizations move only what needs to move and govern everything from one place.
Cloud Strategy Is a Data Governance Problem
The cloud repatriation trend reflects a broader maturation in enterprise IT thinking. Organizations managing this transition well share a common foundation: they governed their data before making infrastructure decisions.
Hybrid is here to stay. The enterprises that bring order to their data first will carry that advantage into every infrastructure decision that follows.
