Why scaling AI and proving ROI are now the real challenge for European organizations.

What comes next is far less straightforward.

For some time, the European AI narrative was fairly comfortable: lots of enthusiasm, plenty of pilots, and just enough regulatory drama to keep things interesting. Companies could experiment broadly, point to a few wins, and call it a strategy.

IDC’s recent research, based on a survey of 200+ European organizations conducted in late 2025, tells a story that is a tad inconvenient for anyone still in “innovation exploration” mode: more than half of European companies report that over 50% of their AI projects have already delivered measurable business outcomes. This is no longer a single pilot result; it is becoming a pattern. And patterns have a tendency to change expectations.

Europe is past the “AI is interesting” phase, but not quite at “AI is effortless” either. Most organizations are somewhere in the messy middle: proof points, momentum, but still unable to explain why that momentum is not turning into something more systematic. Nearly 9 in 10 say their ability to scale AI has improved. And yet, a large portion is operating with what you might call partial discipline. They are moving forward, but without the playbooks, governance structures, and execution models that make scaling feel less like controlled improvisation.

The technology was never the hard part of AI scaling
European organizations are not struggling to build AI. They are struggling to absorb it. When asked what most prevents them from realizing the full potential of their AI investments, the top answers were competition with other transformation priorities, regulatory uncertainty, resistance to process change, difficulty proving ROI, and budget pressure. None of these are technology problems. The blockers are organizational, political, and structural. Throwing more engineering at them will not help.

This is, in fact, a sign of progress. Europe’s AI constraints have shifted from technical feasibility to enterprise commitment, which means the technology has largely done its job. The hard part now is everything surrounding it: sponsorship that survives the next budget cycle, processes redesigned after years of inertia, and ROI demonstrated clearly enough to compete with every other initiative in the budget allocation process. AI is now being tested as a business program, and business programs depend on organizational discipline.

But can organizations measure AI ROI and business impact?

European organizations are no longer just tracking model performance or project completion. Operational efficiency, user adoption, business KPIs, and financial outcomes are all on the scorecard now. This removes a certain flexibility that AI teams might have previously enjoyed. A technically elegant deployment that nobody uses is no longer a qualified success. It is simply not a success.

The encouraging news is that many organizations are starting to respond, with a clear move toward formal business metrics and ROI logic built in from the start.

The gap is widening

Europe’s AI market is entering a separation phase. This is the point where the gap between organizations that can operationalize AI and those still generating isolated use cases starts to widen. The organizations pulling ahead are building the necessary connective tissue: prioritization discipline, outcome measurement, and governance that works at speed. Meanwhile, those still in exploration risk producing impressive narratives about their AI journey while actual business outcomes remain limited.

For enterprise leaders, IDC research is clear about what separates the scalers from the stragglers:

Stop treating AI as a project portfolio. Projects create motion; systems create lasting value.

Build measurement in from day one, not just as good practice, but because organizations that cannot prove value will lose internal budget competition to those that can.

Treat governance as a speed advantage. Organizations that build compliance into reusable controls will move faster, not slower, than those handling it case by case.

For vendors and service providers, the message is equally clear: more features are not the answer to executive skepticism. Proof of business impact is becoming a primary buying criterion. The ability to show how value will be measured, attributed, and reviewed matters more than model benchmarks.

Want to go deeper?
These dynamics are part of a broader shift shaping IT investment across EMEA in 2026. In our upcoming webcast, IDC analysts will explore where growth is materializing, how AI maturity is evolving from pilots to scaled deployment, and what separates organizations that are successfully operationalizing AI from those that are not.

Join us on April 28 for a data-driven view of the trends behind these changes and what they mean for your strategy. Register here.

Ewa Zborowska - Research Director, AI, Europe - IDC

Ewa Zborowska is an experienced technology professional with 25 years of expertise in the European IT industry. Since 2003, she has been a member of the IDC team, based in Warsaw, researching IT services markets. In 2018, she joined the European team with a specific emphasis on cloud and AI. Ewa is currently the lead analyst for IDC’s European Artificial Intelligence Innovations and Strategies CIS.