95% of Dutch organisations are running AI programmes. That makes the Netherlands the top adopter in Europe. But only 5% report real, measurable business value from those programmes.
That gap between deployment and outcome is the most important number in Dutch business technology right now. It tells you that the problem is not access to AI. The problem is knowing what to do with it.
Everyone is running AI. Almost nobody is getting results.
The numbers are hard to argue with. Nearly all large Dutch organisations have some form of AI programme running. Mid-market companies are following: 84% of Dutch SMEs plan to increase AI investment in the next three years. Three million Dutch adults now use AI tools daily, an increase of half a million in just six months.
At the same time, the failure rate is staggering:
- 80% of enterprise AI projects do not meet their business objectives (RAND Corporation, 2025)
- 95% of generative AI pilots fail to reach production (MIT Sloan, 2025)
- 42% of companies abandoned at least one AI initiative in 2025, up from 17% the year before
Companies are spending more, but most are not getting anywhere. The average sunk cost per abandoned AI project is EUR 4.2 million.
The problem is rarely technical.
When organisations fail to generate value from AI, the cause is almost never a broken model or a missing tool. Harvard Business Review put it plainly: the problem is more often organisational and cultural.
Common patterns behind the failures:
- “Success” was never defined before the project started
- AI treated as an IT initiative rather than a business change
- Pilot projects that never had a plan for production
- Data quality issues discovered only after months of development
- C-suite sponsorship that disappears within six months
Only 29% of executives say they can confidently measure AI ROI. In the Netherlands specifically, 85% of organisations lack clear metrics for AI value. If you cannot measure it, you cannot manage it.
What the 5% do differently.
The companies that are seeing actual returns share a few traits. None of them are about having better technology.
They start with the business problem, not the technology. The question is never “how can we use AI?” It is “what operational bottleneck costs us the most time or money, and could AI reduce it?”
They scope tightly. The average failed AI project burns millions before being abandoned. Successful projects have a defined scope, a defined timeline, and a clear owner from day one.
They buy more than they build. MIT research shows that purchasing specialised AI solutions from vendors succeeds roughly 67% of the time. Building from scratch succeeds about one third as often.
They invest in data foundations first. 70% of variation in AI project success traces back to data quality and integration. The companies that win spend time on plumbing before they touch models.
They keep the CEO involved. Projects with sustained C-level sponsorship succeed 68% of the time. Without it, the number drops to 11%.
The mid-market opportunity.
Mid-market Dutch companies (50 to 500 employees) are in an interesting position. They are investing more, but currently report only 35% ROI against a 41% target they would need to consider their AI investments a success.
The upside is clear: firms that get implementation right are seeing 26-31% cost savings in operations, finance, and customer service. Workers using AI effectively save an average of 52 to 60 minutes per day, time they can redirect to strategic and creative work.
But the path matters. Companies without basic digital foundations (automated accounting, centralised data, integrated IT systems) are 43% more likely to abandon AI initiatives within the first year. The Netherlands has strong digital infrastructure: 81% of Dutch SMEs already operate in cloud environments. But cloud access is not the same as data readiness.
The EU AI Act adds urgency.
Full compliance with the EU AI Act is required by August 2026. Prohibited AI practices and AI literacy obligations already apply since February 2025. Fines run up to EUR 35 million or 7% of global annual turnover.
Organisations need to establish a complete AI inventory with risk classification, implement AI literacy training, and prepare technical documentation. For companies that have deployed AI widely without clear governance, this is not a minor compliance task.
Given that 80-90% of Dutch SMEs historically struggled with GDPR compliance, the AI Act represents a significant new challenge. Getting your AI house in order is no longer optional.
What this means for your business.
If your organisation is among the 95% that have deployed AI but are not seeing results, the fix is not more technology. It is a structured, honest assessment of where AI can actually help, what needs to happen first, and whether the expected return justifies the effort.
That is exactly what Practical North’s North Star Audit is built for: three focused hours to map your workflows against current AI capabilities, identify the few opportunities worth acting on, and define a clear next step.
The technology is not the hard part. Knowing where to point it is.