Cloud Repatriation: Can European and Local Cloud Providers Win Workloads Back from Hyperscalers?

Geopolitical uncertainty, rising costs, regulatory pressures, and growing dependence on proprietary technologies are prompting European companies to reassess their relationships with US hyperscalers. Recent research suggests that local and regional cloud providers are becoming increasingly relevant in this debate, raising the question of whether 2026–2027 could mark the beginning of a broader shift towards cloud repatriation to European providers.

This is no longer just a theoretical discussion. According to a Gartner survey published in November 2025, 61% of CIOs and IT leaders in Western Europe expect geopolitical factors to increase their reliance on local or regional cloud providers. Furthermore, 53% anticipate that the same geopolitical considerations will limit their organisations' future use of hyperscale providers.

These findings point to a significant shift in IT infrastructure priorities. After more than a decade of cloud-first strategies, organisations are beginning to question whether concentrating critical applications and data on a handful of global platforms, such as AWS, Microsoft Azure, and Google Cloud, is still the best approach.

Today, cloud infrastructure decisions are increasingly assessed against a broader set of criteria: total cost of ownership (TCO), data sovereignty, regulatory exposure, architectural portability, and long-term operational resilience.

Across Europe, including Romania, GDPR requirements, the implementation of NIS2 and DORA, and wider concerns about technological sovereignty are adding new considerations to cloud procurement and risk management.

Against this backdrop, cloud repatriation and workload diversification are gaining momentum, creating new opportunities for European, regional, and national IT infrastructure and cloud providers.

The question is not whether hyperscalers are becoming irrelevant, but whether their dominance is creating enough economic and strategic reasons for organisations to reconsider where certain cloud workloads should run.

Is Cloud Repatriation a Real Trend or Still an Exception?

Cloud repatriation is becoming an increasingly important topic in enterprise IT infrastructure discussions. However, it should not be mistaken for a widespread retreat from the public cloud. Organisations may move certain applications from hyperscale environments to private clouds, dedicated infrastructure, or colocation services while continuing to expand their use of public cloud services for other workloads.

The reasons vary from one organisation to another. Some businesses are responding to higher-than-expected operating costs, others to regulatory or data sovereignty requirements, and others to the risks associated with dependence on proprietary technologies.

Gartner's findings suggest that geopolitical factors are becoming particularly important in Europe. The same survey found that 55% of CIOs and IT leaders identified open-source technologies, such as OpenStack, as an important factor in their future cloud strategies.

This is a strong indication that open technologies and infrastructure portability can give organisations greater flexibility to switch providers or redistribute workloads, even when there are no immediate migration plans.

Cloud repatriation should therefore be seen as a reassessment of how IT resources are allocated, rather than a reversal of digital transformation.

For example, a company might move a high-traffic database or a business application with predictable resource requirements to dedicated infrastructure while continuing to use hyperscale platforms for AI, advanced analytics, or globally distributed services.

The key question is no longer whether the public cloud is inherently better than private or regional infrastructure. Instead, it is whether a particular workload benefits enough from hyperscaler capabilities to justify the associated costs, complexity, and dependencies.

Rising Hyperscaler Costs Are Driving the Conversation

One of the main arguments for moving from hyperscalers to European or national cloud providers is the potential to reduce total infrastructure costs without giving up the benefits of the cloud model.

AWS, Microsoft Azure, and Google Cloud offer extensive service ecosystems and scalability that few providers can match. However, these advantages come with complex pricing models, where the actual cost of running an application can significantly exceed initial estimates.

Beyond compute resources, cloud bills may include storage, provisioned IOPS, snapshots, backups, load balancing, networking services, monitoring, inter-region data transfers, and outbound traffic.

For many enterprise workloads, particularly those with relatively stable resource consumption and clearly defined performance requirements, European and regional cloud providers can offer a competitive alternative through simpler pricing models, more predictable costs, and greater contractual flexibility.

However, comparisons should be based on total cost of ownership (TCO), not simply the price of CPU, RAM, or storage. A cheaper virtual machine does not automatically translate into genuine savings.

An accurate assessment must cover all the services required to run the application, availability and security levels, technical support, migration and integration costs, and the infrastructure's ability to meet future requirements.

Over a three-to-five-year period, differences between commercial models can become substantial, particularly for applications running continuously and handling large volumes of data.

Moving workloads to European or local cloud providers is therefore not just about cutting costs. It is also about improving IT budget predictability and gaining greater control over long-term spending.

Egress Costs: When Data Transfers Become a Significant Business Expense

One of the less visible aspects of cloud economics is the cost of transferring data out of a provider's infrastructure.

In a previous M247 Global article, Egress Cost in IoT Projects: The Hidden Tax of Cloud-Centric Architectures, we examined how outbound data transfer charges can fundamentally change the economics of applications processing large volumes of data.

While inbound traffic (ingress) is generally free, hyperscale providers typically charge for outbound data transfers (egress), with rates varying by region, volume, destination, and service.

Pricing differences between providers can be considerable.

Although these figures are based on pricing published on the providers' websites, they are indicative and should not be interpreted as universal rates or directly comparable offers. Actual costs depend on the region, contractual terms, service configuration, included data transfer allowances, and traffic volumes.

Even when per-gigabyte charges appear relatively small, they can add up quickly for applications involving continuous data exchanges or multiple integrations with external systems.

In an IoT architecture, for example, data may be collected from thousands of devices, processed centrally, and then distributed to operational applications, analytics platforms, and third-party systems. Each additional destination can increase the volume of billable outbound traffic.

Hybrid and multicloud environments introduce another layer of complexity. Moving data from one cloud provider to another may incur egress charges at the source, along with additional connectivity, processing, or storage costs at the destination.

These charges can influence architectural decisions, encouraging organisations to keep applications and datasets within a single provider's ecosystem, even when other environments could offer better economics.

Vendor Lock-In: The Cost of Architectural Dependency

Financial considerations are only part of the picture. Over time, many businesses have built increasingly sophisticated application architectures around hyperscaler-specific services, including proprietary databases, event-processing platforms, serverless environments, identity systems, and managed data analytics services.

These capabilities accelerate development and reduce operational overhead. At the same time, they can increase the cost and complexity of switching infrastructure providers.

Migrating a conventional virtual machine is relatively straightforward compared with redesigning an application that depends on multiple proprietary services.

The challenge is not necessarily application code portability, but the portability of the entire operating environment: data models, orchestration, identity management, networking, monitoring, automation, and operational procedures.

Containerisation, Kubernetes, infrastructure as code, and open technologies can improve portability, but they do not automatically eliminate dependencies. Managed Kubernetes implementations, proprietary networking components, and cloud-specific storage integrations can still create significant migration challenges.

Many CIOs are already asking whether this level of architectural dependency is an acceptable trade-off for the productivity and innovation benefits offered by a particular cloud ecosystem.

Growing interest in open standards, interoperability, and infrastructure portability suggests that organisations want to keep their future options open, even when they have no immediate plans to leave their current providers.

Digital Sovereignty Is Becoming a Business Requirement

In Europe, cloud infrastructure decisions are increasingly influenced by geopolitical uncertainty, regulatory obligations, and concerns about technological dependency.

Data residency has long been important for regulated industries, but today's discussion goes well beyond where information is physically stored.

European organisations are paying closer attention to who operates their infrastructure, which legal jurisdictions apply, how administrative access is controlled, and whether critical services could continue operating under adverse geopolitical or commercial conditions.

Hosting data within the European Union does not, on its own, resolve all these concerns.

A data centre located in the EU but operated by a non-European provider may offer strong contractual and technical safeguards. Nevertheless, customers still need to assess the provider's ownership structure, legal exposure, operational dependencies, and applicable protection mechanisms.

Similarly, choosing a European provider does not automatically guarantee compliance or eliminate cybersecurity risks. Under NIS2, for example, an important consideration is an organisation's ability to demonstrate appropriate risk management and control.

For organisations subject to GDPR, NIS2, DORA, or other sector-specific regulations, cloud procurement decisions increasingly require collaboration between IT architecture, security, compliance, legal, and procurement teams.

In this context, regional providers may offer several advantages: greater contractual flexibility, local operational accountability, direct access to technical teams, and better visibility into where and how workloads are operated.

Whether these advantages are decisive depends on the application, the regulatory environment, and the provider's actual capabilities.

Can National and Regional Cloud Providers Compete?

For European and national infrastructure providers, cloud repatriation presents a potential growth opportunity.

However, winning workloads back from hyperscalers requires more than competitive pricing or available capacity in local data centres.

Enterprise customers expect operational maturity, documented service levels, robust security controls, reliable connectivity, disaster recovery options, and integration with their existing IT environments.

Migration support is equally important. The business case for switching providers can quickly weaken if applications require extensive redesign, prolonged downtime, or significant internal engineering resources.

Regional providers must therefore demonstrate that they can deliver a reliable and mature operating model, rather than simply offering an alternative hosting location.

This is precisely where European providers such as M247 Global can differentiate themselves: by combining infrastructure services with networking expertise, high-speed global networks, predictable commercial terms, responsive technical support, and architectures that support hybrid environments.

This is particularly relevant for markets such as Romania, where locally hosted infrastructure can serve both domestic organisations and international businesses operating across Central and Eastern Europe.

So, can European and national cloud providers compete directly with the likes of Amazon, Google, and Microsoft?

Yes. By offering competitive technologies, proximity to end users, strong regional connectivity, operational transparency, and the ability to negotiate infrastructure requirements directly, European and local providers can deliver tangible advantages to enterprise customers.