Even though 95% of AI initiatives deliver zero value, companies feel enormous pressure to ‘do something with AI’ – now. The fear of disruption is causing leaders to jump straight to solutions without asking the fundamental questions: Do we have the data we need? Does this solve a real customer problem?
The result is that many companies are investing in AI tools that can only analyse publicly available data. Their competitors are doing the same, which is why very few are creating the competitive advantage they hoped for.
This is one of the main points from INSEAD’s webinar ‘Making AI transformation work: Avoiding the mistakes from Digital Transformation’, where Professor Nathan Furr and transformation consultant Steven Koolen share experiences from the successful digital transformation of the Netherlands Lottery. These lessons are more relevant than ever in the AI era.
The solution: People control, AI supports
A typical mistake is for companies to embark on AI projects driven by fear; this could be the fear of becoming irrelevant, of competitors overtaking them, or of their business model becoming obsolete. Value is created when you start at the opposite end: the customer’s problem and the employees’ everyday life.
Nathan Furr cites an early example of AI involving chess legend Garry Kasparov, who was defeated by IBM’s Deep Blue. Most people have heard this story. But a lesser-known angle is that Kasparov subsequently became very interested in AI and early predicted the following:
The future is not about replacing humans with AI, but about humans becoming ‘shepherds’ – that is, shepherds who lead flocks of AI agents. This is already being put into practice at leading companies, where technical support staff receive AI suggestions in real time during customer calls. The purpose is not to take over the conversation, but to make the employee better, faster and more relevant.
The result is both lower costs and higher customer satisfaction, because human judgement still controls, while AI handles the administrative tasks and suggests solutions.
The case: Nederlandenes Loterij – from internal focus to customer value
Nederlanden Loterij found itself in a situation that many established companies will recognise: fragmented legacy IT systems, a limited budget and sudden competition from digital players without the technical debt. Customers moved from physical stores to online. The new – and purely digital – gaming companies gained market share.
Start with ‘why’ – not the technology
Nederlanden Loterij did something fundamentally different: they didn’t start with the technology, but with the purpose. The company exists to support sport and society – and to ensure that people who want to gamble do so in a safe place without the risk of gambling addiction. The digital transformation was not sold as ‘we need new technology,’ but as ‘this is how we protect our mission and relevance in a digital world.’
The challenge was paradoxical: the company was doing well financially, which created a certain complacency within the organisation. But management saw the long-term risks: without action, they would lose relevance as customers moved online and digital competitors gained ground. The key was to make the need concrete on two levels: firstly, to show the strategic threat, and secondly, to demonstrate how transformation would remove frustrating internal processes and give employees more time for customers.
Transformation on three levels: Everyone must be involved
Success was based on reaching all layers of the organisation simultaneously:
- Management level (top-down): It sounds trivial, but it proved crucial: Can management tell the same story? When Koolen asked the executive board to present the vision to each other, they discovered that they did not actually agree on the details. Without a common, coherent vision from the top, it is impossible to stay the course through a long transformation. Management must not only believe in it. They must be able to articulate exactly the same ‘why’ and ‘where to’.
- Employee level (middle management): Here, it was about demonstrating concrete value in everyday life. Not abstract goals about digital maturity, but tangible improvements: Fewer internal meetings. Less silo thinking. More time for customers. Employees were already frustrated about being too internally focused – the transformation gave them the tools to change that. It created motivation because people could see how their working lives were improving.
- Bottom-up: The employees who are closest to the customers often know best what needs to be changed in practical terms. Management can set the direction, but it is impossible for them to know all the details. Therefore, front-line employees were actively involved in identifying problems and designing solutions. This ensured both better solutions and ownership of the change among those who had to execute it.
The balance between control and empowerment
The key was to combine top-down direction with bottom-up innovation. Too much top-down will result in solutions that do not fit reality. Too much bottom-up, and the effort will be fragmented in a hundred different directions without a common goal. Nederlandenes Loterij found the balance by letting management define the ‘why’ and “what” (purpose and business goals), while teams were given the freedom to figure out the ‘how’ within their areas.
Gain insights so that AI creates business value
As a long-standing partner of INSEAD for more than a decade, we at the Scandinavian Executive Institute are interested in how technology – such as AI – influences the role of leadership. AI is gradually becoming more prominent in both our executive education programmes and board education programmes, but always with a focus on how leaders can use new knowledge and the experiences of others to create business value.
The experiences of Nederlandenes Loterij show the way: Successful AI transformation is not about technology first, but about starting with purpose, involving the entire organisation, and using AI to make people better – not replace them. It takes courage to say no to quick fixes and yes to the slower but more sustainable path through digital maturity first, AI value second.
