By Kevin Cochrane
European businesses are increasingly reassessing reliance on hyperscalers, exploring how alternative cloud providers can support data sovereignty, pricing transparency, and workload-specific infrastructure needs.
Are European businesses rethinking their cloud strategy? As data sovereignty, AI performance demands, and cloud economics reshape the market, organisations are increasingly looking beyond hyperscaler defaults and embracing alternative cloud providers. In this article, Kevin Cochrane, CMO at Vultr, explores why cloud selection is becoming a strategic business decision rather than an automatic choice.
Traditionally, hyperscalers have been the default for European businesses, offering the path of least resistance to the cloud. Today, however, many businesses are beginning to question whether this is the right approach after all. European businesses are migrating workloads to alternative cloud providers, drawn by greater control, competitive commercial terms, and flexible infrastructure able to be built around specific workloads.
Cloud infrastructure enters a new phase of maturity
Putting all your eggs in one basket is risky, but until recently, this is exactly what European businesses did, entrusting one of the major hyperscalers with their entire infrastructure stack.
It’s understandable why. For much of the last decade, hyperscalers offered the simplest route to cloud adoption. They provided scalable, standardised infrastructure that reduced friction for organisations moving away from on-premise systems. This ease of adoption created a strong default position across industries.
That default position is now weakening. While hyperscalers continue to dominate market share – accounting for 63% of the global enterprise cloud infrastructure services market – European businesses are increasingly reassessing whether default reliance on a small number of providers remains the right long-term strategy.
With more than 61% of IT leaders in western Europe planning to increase their reliance on local and regional cloud providers, cloud infrastructure is clearly entering a new era.
The data sovereignty imperative
Much of this shift is being driven by data sovereignty requirements. Due to rules including the EU’s Digital Operational Resilience Act (DORA) and the NIS2 Directive, regulated businesses must now guarantee transparency and auditability across their digital supply chains. This requirement demands greater insight into and control over their cloud workloads.
Meanwhile, the US CLOUD Act has made the handling of cloud workloads a red hot topic. Under the law, US hyperscalers can be legally compelled by the US government to hand over data, even if the data is stored in a European city.
The EU has hit back through its Data Act, which prohibits unlawful third-country access to non-personal industrial data. In the UK, meanwhile, the government is pushing forward a Cyber Security and Resilience Bill, which will mandate data sovereignty as a non-negotiable for critical national infrastructure.
In the midst of these new rules and regulations, European businesses increasingly see the value of local, natively sovereign cloud alternatives. Figures from Accenture suggest that as many as 62% of European organisations are currently seeking sovereign solutions.
Alternative hyperscalers can help enterprises meet this goal. Leading alternative cloud providers offer isolated, country-specific environments that feature dedicated control planes and strict data residency controls. These features ensure that a business’s data, processing power, and governance remain entirely confined within national borders.
Pricing and workload considerations
Data sovereignty aside, European businesses have two other very good reasons to look to alternative cloud providers. The first is the need for more transparent and competitive pricing models.
An investigation by the UK’s Competition and Markets Authority (CMA) found that the hyperscaler market is highly concentrated, and that less than 1% of customers switch providers or use multiple clouds due to lock-in resulting from either poor interoperability with other providers or from commercial barriers. The CMA is proposing to impose targeted and bespoke interventions to help open the market.
The European Commission is following suit by launching a sector-wide antitrust probe into hyperscale infrastructure, largely driven by complaints from startups and enterprises regarding hyperscalers’ opaque billing practices.
As organisations adapt to the AI boom, many will be looking for a way out of these inflexible and pricey contracts. Alternative hyperscalers offer that solution, with many providing transparent pricing plans that remove the cost, complexity, and hidden data-extraction fees associated with traditional cloud providers.
AI will help drive change in one other way. As AI workloads make latency and performance increasingly important, one-size-fits-all cloud strategies are no longer fit for purpose. Increasingly, cloud decisions will be made at the workload level rather than the platform level, with infrastructure selected based on its ability to deliver the right price-to-performance metric.
The right alternative cloud provider can address this requirement through full-stack, cloud-native AI and GPU infrastructure distributed across an extensive network of regional data hubs, allowing enterprises to run highly demanding workloads locally without sacrificing scalability.
The exodus from the hyperscalers
The impact of all these pressures are showing up in the data. In the UK, what were once quiet rumblings about migrating to alternative cloud infrastructure have become everyday conversations. The effect could be huge. According to one study, 87% of UK businesses are planning to repatriate some, if not all, of their workloads over the coming two years.
In some regions, this transition is being driven by regional policies. Spain, for example, has mirrored the successful approach it took to its nationwide green energy deployment. By establishing lighter regulatory frameworks, Spain’s government has turned the country into a magnet for alternative cloud infrastructure providers.
Meanwhile, the workload-focused mindset is driving dramatic change in Germany, where organisations are looking beyond generalised cloud infrastructure intended for office productivity and instead focusing spend on the area where the workload demands it most: heavy-duty AI for industrial robotics, simulation, and manufacturing.
Research from Bitkom revealed that 42% of German industrial firms have already integrated AI directly onto the factory floor, with 32% using it for complex predictive maintenance and nearly half (48%) deploying compute-heavy Digital Twins to simulate entire production lines.
Specialised AI workloads are driving specialised infrastructure partnerships. Mistral, a French AI company, is a prime example. The company bypassed traditional single-provider lock-in by designing its models to scale globally through targeted partnerships with alternative, high-performance cloud platforms. It has access to all the local infrastructure it needs for control and performance, but it also has a foundation for global reach.
From default selection to strategic choice
Cloud infrastructure is becoming an intentional strategic choice. Hyperscalers are no longer automatically the starting point. European businesses are increasingly selecting infrastructure based on the needs of individual workloads and business outcomes, rather than defaulting to a single provider.
This shift reflects a broader maturation of the cloud market. As regulatory requirements evolve, scrutiny of cloud economics intensifies, and AI creates new performance demands, organisations are reassessing long-held assumptions about where and how workloads should run.
The result will be a diversified infrastructure that’s more open, less expensive, and more compliant. The question is no longer whether to use the cloud, but how to use it most effectively. In this environment, cloud selection becomes less about default adoption and more about strategic design.
Kevin Cochrane
