Public Cloud Development And Infrastructure Platforms In Europe

Analyst research rarely describes buyers as being strong-armed. Forrester's commentary on this market does.

Its assessment is that some client organisations may be swayed or strong-armed by their local regulators and politicians into looking at alternatives to the hyperscalers, and that sovereignty requirements have the potential to disrupt the European market.

That sentence identifies what makes this evaluation different from every other cloud comparison. In most markets the decision is technical and commercial: which platform has the services you need at a price you can justify. In Europe a third force applies, and it originates outside the IT department.

Thirteen providers, six European

The Forrester Wave: Public Cloud Development And Infrastructure Platforms In Europe, Q1 2023 evaluated thirteen providers against thirty criteria: Amazon Web Services, CloudSigma, Google Cloud, IBM, IONOS Cloud, Microsoft, Oracle, Orange Business, OVHcloud, Salesforce, SAP, T-Systems, and Vultr.

Six of those are European: CloudSigma in Switzerland, IONOS and T-Systems in Germany, Orange Business and OVHcloud in France, and SAP in Germany.

Compare that against Forrester's global evaluation from Q4 2022, which scored ten vendors against thirty three criteria and named Microsoft one of only two Leaders. The global list is a hyperscaler contest. The European list is a hyperscaler contest with six additional companies whose primary argument is not capability.

Orange Business placed as a Strong Performer, standing out on strategy with high marks for product vision and strong results for its core services portfolio.

The trade the evaluation describes

Forrester's summary of the result is direct: the European landscape is dominated by hyperscalers, which shine in core competencies such as storage and compute and in the ability to scale rapidly and build resilience into their services.

That is a plain statement that the largest providers are better at the things cloud platforms are fundamentally for.

But European-headquartered organisations rated as Strong Performers made strong claims in data sovereignty requirements and sustainability, and Forrester notes that in very specific use cases some Strong Performers excelled over the Leaders in AI and machine learning.

So the trade is legible. Breadth, depth, scale, and pace of innovation on one side. Sovereignty, sustainability positioning, and occasional specialist advantage on the other.

Which side matters depends on something the evaluation cannot score, because it depends on whether your organisation faces a requirement that capability does not satisfy.

What sovereignty actually means

The word gets used loosely, and it covers at least three distinct requirements with very different implications.

Data residency is the weakest form. Data is stored in a specified jurisdiction. Every major provider offers this and it has been solved for years.

Operational sovereignty is stronger. It concerns who can access the data and the systems, from where, under whose employment, and under which legal authority. A platform operated by staff outside the region, or by an entity subject to foreign disclosure obligations, satisfies residency and not this.

Technical sovereignty is strongest. It concerns whether the organisation could continue operating if the provider withdrew, whether the technology stack is controlled within the jurisdiction, and whether the encryption keys are held somewhere the provider cannot reach.

Most sovereignty conversations conflate these, and vendors are content to let them. A provider answering a sovereignty question with a data centre map has answered the first and neither of the others.

That gradation explains why the European providers in this evaluation compete credibly despite the capability gap. For an organisation whose requirement stops at residency, the hyperscalers are fine. For one whose requirement reaches operational or technical sovereignty, the question changes shape entirely.

Forrester notes that hyperscalers have been doubling down on cloud sovereignty initiatives and working with regulators to demonstrate transparency about where activities are processed and how customers are supported in the region. That is a real response and it has advanced considerably since this evaluation, which is worth checking against current offerings rather than against 2023 positions.

The customer finding nobody advertises

Forrester published a companion trend report from customer interviews, and its title captures the demand drivers: cost reduction and sovereignty concerns drive cloud adoption in Europe.

The trends it identifies are multivendor strategies, increasing adoption in regulated industries, and difficulties in reaping benefits from implementations.

That third one is the honest finding and it deserves more attention than it gets.

Cloud programmes routinely underdeliver against their business cases, and the reasons are consistent across geographies. Applications migrated without modification cost more to run on metered infrastructure than they did on owned hardware. Organisations pay for capability they never adopt. Cost management is nobody's job until the invoice becomes a problem. And the operating model that made sense for a data centre does not translate.

The European version adds a specific complication. Forrester notes that multicloud in Europe carries one further layer of complication compared to other geographies, and the reason is that a multivendor strategy driven by sovereignty is not the same as one driven by capability.

A capability-driven multicloud deployment puts workloads where they run best. A sovereignty-driven one puts workloads where they are permitted, which frequently means running the same class of workload on two platforms with different tooling, different skills, and different operational models, because some of it is subject to a requirement and some is not.

That is expensive, and it is a cost that appears in operations rather than in the licence comparison used to justify the decision.

What has changed since 2023

This evaluation is three years old and the sovereignty environment has moved considerably, which affects how much weight to place on the specific results.

European legislative activity around data access, cloud switching, and interoperability has advanced, with provisions aimed at reducing switching barriers and clarifying obligations around non-personal data. Sovereign cloud offerings from the hyperscalers have expanded substantially, including arrangements involving local operating partners and jurisdictionally separated control planes.

The geopolitical backdrop has also shifted in ways that make the political dimension Forrester identified more prominent rather than less. Public sector procurement in several member states has become more explicit about sovereignty requirements, and the debate about dependency on non-European infrastructure has moved from specialist forums into mainstream policy discussion.

None of that changes the structural trade. It changes where the line falls, and it has generally moved in favour of asking harder questions.

The practical consequence is that anyone using the 2023 evaluation should treat the vendor set and the criteria as sound and verify current sovereignty offerings directly, because this is the fastest-moving dimension in the market and vendor positions from three years ago do not describe today's.

The political dimension, stated carefully

Forrester's language about clients being swayed or strong-armed acknowledges something that most technology research avoids: that some of the pressure in this market is political rather than regulatory.

Regulation creates obligations that can be read, assessed, and complied with. Political preference creates expectations that are harder to specify and equally consequential, particularly for organisations in public sector supply chains, regulated industries, or sectors designated as critical.

An organisation can satisfy every applicable regulation using a non-European provider and still face questions from a regulator, a minister, a works council, or a customer about why it made that choice.

There is a reasonable case on each side of this and it is not a technology argument. Proponents of European provision point to strategic autonomy, resilience against extraterritorial legal reach, and the value of a domestic industry. Proponents of the hyperscalers point to capability, cost, innovation pace, and the risk of building on smaller platforms with less investment behind them.

What a buyer can usefully do is separate the questions. Establish which requirements are legally binding, which are contractually imposed by customers, and which are preferences held by stakeholders. Those three have different weights and different negotiating positions, and treating a preference as an obligation removes options unnecessarily while treating an obligation as a preference creates exposure.

Where this leaves the decision

For most European organisations the honest position is that the hyperscalers remain technically superior for general workloads, and that the European providers are credible for the subset of workloads where sovereignty requirements bite.

That produces the multivendor pattern Forrester's customer research identified, and it is a rational response to genuinely different requirements rather than indecision.

What it requires is deliberate placement. Deciding which workloads carry sovereignty requirements, which do not, and running each where it belongs, with the operational cost of that split acknowledged in the business case rather than discovered afterwards.

The alternative pattern, where an organisation applies its strictest sovereignty requirement uniformly across everything, is common and expensive. It happens because nobody wants to be the person who classified a workload as unrestricted and turned out to be wrong, which is an entirely understandable incentive and a costly default.

Making that classification explicit, documented, and owned by someone with the authority to defend it is the single decision that determines whether a European cloud strategy is coherent or merely cautious.

Forrester does not endorse any vendor named here, and tier placement should not be read as a recommendation to buy. Regulatory and policy positions described here change frequently and should be verified against current sources.