The Real Constraint on Innovation: Q&A with Zbigniew Lewicki, Magnum Ice Cream Co.

Zbigniew Lewicki is Chief Research, Design & Innovation Officer at the Magnum Ice Cream Company, where he leads global innovation strategy across product development, design, and R&D. His work focuses on how organizations translate insight into scalable innovation, combining human expertise, data, and emerging technologies to drive category leadership. He brings expertise in innovation strategy, product design, R&D, AI-enabled transformation, and the culture and decision-making systems that turn insight into action.

IQ: Many organizations have an innovation strategy, yet pipelines are crowded and slow. Is the breakdown in ideation, development, or scaling?

Mr. Lewicki: I would reframe the question, because the starting point is not always as strong as we assume. Many companies have a strategy, but fewer have one that genuinely expands the category. Too often, it defaults to a safer place—renovation and optimization of the existing portfolio—rather than the true category leadership where growth innovation should begin. And even with a strong strategy, the breakdown does not sit within ideation, development, or scaling. It sits much deeper, within culture. Innovation has to be in the company’s DNA, visible in how leaders behave and what is incentivized. The issue is not process or organizational design. It is fundamentally culture.

IQ: With so much data and AI-driven insight, leaders are seeing more signals than ever about shifting consumer behavior. How do you distinguish what truly matters from noise?

Mr. Lewicki: If you look at consumer goods and food in particular, the pace of change over the last 12 to 18 months has been extraordinary. You have developments like GLP-1, changes in how consumers think about nutrition, such as the reversal of the food pyramid, new regulations on sugar and HFSS in markets like the UK, and consumer-facing applications like Yuka reaching tens of millions of users. These are tectonic shifts, and the speed at which they are happening is comparable to what we might have seen over a decade in the past.

We are also beginning to see new behaviors emerge, including early forms of AI-driven purchasing where systems or agents start to assemble shopping baskets on behalf of consumers.

To distinguish signal from noise, you need a structural solution. It starts with clear strategic imperatives that are informed by data and supported by AI. Increasingly, companies should be thinking about how AI agents will make decisions for consumers and which attributes of their products will make them the preferred choice in that environment. The answer is not to consume more data. The answer is to build an innovation engine that translates macro trends into clear, actionable criteria for decision-making.

IQ: Where are you seeing AI genuinely change the speed and quality of innovation today—and where is it still falling short?

Mr. Lewicki: AI is both a buzzword and a real disruptor, and it is a positive disruptor in many ways. But one of the limitations is that we often look at it in isolation rather than as part of a broader system. I think about what I would call an intelligent innovation lab. This is an environment where you combine the best of AI models with human expertise and situate that within a modern, automated, or robotic workplace.

AI can do a very strong job of triangulating consumer data with product composition and other inputs. But it cannot fully capture the nuance of sensory experience, such as taste, texture, or the emotional component of a product. That is why human expertise remains essential. The future of innovation is not about AI replacing people. It is about AI and humans co-creating within a connected system that includes both digital models and physical, automated environments.

IQ: As AI becomes more embedded, where does human judgment remain essential, and what should be handed over to the machine?

Mr. Lewicki: There is a misconception that creativity is the primary driver of innovation. Pattern recognition is the most powerful driver of innovation effectiveness, and this is an area where AI models perform exceptionally well. They can process large amounts of data and identify relationships that would be difficult for humans to see. At the same time, AI cannot operate in isolation. It requires a human component to guide it, to interpret outputs, and to continuously refine and optimize the models. There needs to be a hybrid learning loop between human expertise and machine capability.

If organizations rely too heavily on AI, there is a risk of losing critical human skills and expertise. Over time, that could weaken the organization’s ability to differentiate itself. So the objective is not to replace human judgment—which remains essential in shaping, validating, and evolving the system—but to find the right equilibrium.