Technology alone does not scale – organizations do
Given these numbers – where are companies going wrong?
We consistently see four root causes: First, companies lack clear decision-making structures. There is no defined ownership for who makes which decisions and who is accountable. Second, there is a skills gap combined with a lack of trust in data. Third, what works in a small, agile team does not automatically translate to end-to-end processes at scale. And fourth, many organizations simply lack the organizational maturity needed. There is no alignment between strategy, governance, skills, and culture.
What does it take for a company to implement AI successfully?
The first question to answer is: what does ‘success’ mean for your organization? That definition must be anchored in business strategy. From there, you need a clear governance and operating model with explicit decision rights across the entire AI portfolio. Processes must be designed end-to-end, from development through to operations. Skills and capabilities need to be built and embedded systematically. And the culture must be aligned: the mindset has to be right.
Only then should you turn to technology and platform decisions. That is the core message: technology is the enabler, not the driver. Organizations that select the technology first and then try to build the organization around it are getting the sequence fundamentally wrong.
The human factor as a reason for project failure has been on the agenda for 30 years. Why are companies still reluctant to invest in change management, even though it is one of the most powerful levers for successful implementation?
Because change management is still treated in many organizations as a communications add-on, not as a core implementation lever. As long as change is seen as a “nice to have,” budgets flow first to technology, process design, and external implementation partners.
Yet in Data & AI initiatives, value is not determined solely by the quality of the solution. What matters equally – perhaps more – is whether people understand it, adopt it, and integrate it into their daily work. That is why change management is not a supplement: it is part of the value chain. Organizations that invest too late risk that even strong concepts never reach real-world adoption.
Why aren’t the numbers convincing leadership to act?
I believe there is still a widespread misconception that change management cannot be measured. In fact, there are clear KPIs that make CM outcomes quantifiable. On top of that, discipline is often perceived as ‘soft, and therefore less relevant than technology or process work.
We were able to demonstrate the impact of change management unambiguously with one client: decision-making time on key governance issues dropped from three to four weeks down to five to seven business days – a reduction of approximately 70%. And the time to first measurable value in Data & AI use cases shortened by 30 to 40%. These are not soft targets. These are hard business outcomes. And that is what changed the conversation.
How can companies overcome employee resistance to AI?
Resistance to new initiatives is entirely normal, and often a signal that the topic is being taken seriously. What matters is how organizations respond to it.
It requires early involvement of employees, a clear change story, champions at multiple levels of the organization, and an iterative feedback mechanism. We recommend a responsive change management approach that starts in the very first project phase, not just before go-live.
People need to understand why AI is being introduced, what will change for them personally, and what concrete benefit they will gain. Only when uncertainty gives way to clarity can genuine acceptance emerge.
What role does leadership play in AI projects, and what has changed compared to other transformation initiatives?
AI must be strategically embedded in the organization and visibly championed at the leadership level. Senior management commitment is therefore not optional; it is a prerequisite. Without aligned incentives, adapted decision processes, and an AI-ready culture, even excellent pilots remain isolated one-offs.
Leaders must act as active sponsors and position themselves as visible role models, through leadership dialogues or other formats that provide orientation and direction.
There is a second critical lever: communication. How AI and its strategic relevance are positioned internally is decisive. Employees need to understand the ‘why’ behind the initiative – not as a top-down directive, but as a shared strategic conviction.
What is the single biggest problem you see in AI projects?
The biggest problem is often the assumption: ‘We can handle this ourselves; it’s just another transformation project.’ Many companies underestimate that AI is not an isolated technology initiative but a fundamental organizational change. That realization often comes only after significant resources have already been invested, without lasting impact.
In practice, I see three recurring patterns: First, lack of ownership – no one feels genuinely responsible for cross-domain data quality. Second, governance is designed as a brake rather than an enabler: Too many policies, too few pragmatic decision frameworks. Third, enablement comes too late. Skills are only built when use cases are already going live. At that point, ambition quickly turns into overload.
What ROI can companies realistically expect?
There is no universal ROI figure. It depends heavily on organizational maturity, the use cases selected, and the depth of implementation. That said, in practice we consistently see significant efficiency gains, faster scaling of relevant use cases, and meaningfully higher active use of data and AI across business functions.
But the real lever is not just short-term savings. It is building an organization that can generate data-driven value in a sustainable and scalable way – over time.
