The AI challenge is fundamentally a data challenge
AI is no longer a technology question—it is a business discipline. Yet most initiatives still start with platforms instead of purpose. To unlock real value from AI, organizations must rethink where they begin: with data they trust, problems worth solving, and value that is defined before a single model is deployed.
At NOVA Day, Ørjan Clausen reminded us why so many AI initiatives fall short. Not because the technology isn’t ready, but because solutions are built on fragile data foundations, poorly defined problems, and value that is never properly anchored in the business.
If AI is to create real, lasting business value, we must start somewhere else than with the tools. We must start with data.
The numbers are uncomfortable—but necessary to acknowledge. Only 4 percent of Norwegian companies say that artificial intelligence is critical to their business model. At the same time, international studies show that up to 95 percent of AI projects never make it into production. The ambitions are there. The budgets are there. The technology is there. Still, the initiatives are quietly abandoned.
The message from the stage was not that AI is overhyped. It was more fundamental than that: technology is rarely the real problem. How we approach the problem is.
Individual Productivity Does Not Equal Business Value
Many organizations are currently seeing increased individual productivity through AI. Research Ørjan referred to shows that nearly half of all AI usage is tailored to individual employees. But a human plus a machine usually just creates a more efficient individual—not necessarily a more effective organization.
Only around one third use AI collectively across the organization, and among these, few consider the technology to be business‑critical. Only a small minority have succeeded in integrating people, processes, and data in a way that fundamentally changes how the organization operates.
Meanwhile, competitors may be gaining momentum. The fear of falling behind is real—but it often leads to action without direction.
We Start in the Wrong Place
The paradox is clear: most AI projects do not fail because of a lack of technology. Platforms, models, and infrastructure are widely available. Many organizations also have data—but they lack control, structure, and a shared understanding of what that data should actually be used for.
Too many initiatives start with the solution. With the tool. With the technology someone has seen in a demo or read about in a case study. The questions that should have been asked first come too late—or not at all:
- What problem are we trying to solve?
- What value should this create?
- How will we know if we succeed?
When these questions remain unanswered, it is hardly surprising that projects come to a halt.
Three Prerequisites for Success
In his keynote, Ørjan pointed to three fundamental prerequisites for realizing value from AI—presented in the right order.
1. Data as the Raw Material
AI is only as good as the data it is built on. Generic solutions may deliver short‑term efficiency gains, but rarely sustainable competitive advantage. The real value lies in proprietary data—data the organization controls, understands, and can combine in new ways. This requires high data quality, consistent values, and clear definitions. Without this, the old rule still applies: garbage in, garbage out.
2. Design as a Process
Design is not about decoration or bureaucracy. It is about forcing clarity before building. Through hypotheses, early testing, and validation, risk is dramatically reduced. A structured design process ensures that you start with the problem—not the solution—and avoid investing in technology that no one actually needs or uses.
3. Value as the Goal
Value cannot be something you assess at the end of a project. It must be defined upfront, measured before and after, and embedded in roles and processes. Efficiency makes us faster in today’s processes. Prediction, learning, and collaboration between people and systems prepare us for tomorrow. Without a clear, shared understanding of what value means in practice, AI becomes a side project—not part of the business model.
Technology Is Never the Strategy
A recurring point in the keynote was that AI in itself is never a strategy. It is a means. Organizations that succeed do not start by choosing platforms. They start by defining business objectives. They build organizations where people and machines work together in end‑to‑end processes, supported by high‑quality data and clear ownership.
For NOVA Consulting Group, this is at the core of how we work. Data first means addressing the hard questions early. Understanding the problem before proposing a solution. Ensuring that investments in technology result in documented, tangible value creation.
Perhaps this is where the real opportunity lies. Not in being first to chase the next technology—but in starting in the right place. With data as the foundation, design as direction, and value as the guiding principle.
Continue the conversation
Ørjan Clausen is available for keynotes to explore how data, design, and business value must come together to make AI work in practice and what it takes to move from experimentation to real impact. Reach out to us if you are interested and want to continue the conversation.
