German retail and tourism cooperative REWE operates thousands of supermarkets and consumer stores, as well as travel agencies, tour operators, and hotel brands across Europe. In this conversation with McKinsey’s Sebastian Gatzer and Silvia Andre, REWE’s chief digital and technology officer Christoph Eltze discusses the need to act early on AI and why strengthening governance, skills, and leadership behavior is as important as managing customer-facing risks. The interview has been edited for clarity and length.
Ultimately, winning with AI is above all a question of culture, teams, and ways of working. We have become a more data-driven company and that cultural asset is something competitors cannot simply buy.
Q: How is AI reshaping your company’s strategy and competitive advantage today?
Christoph Eltze: First, it is important to clarify that when we say “AI” we are encompassing everything that we used to call analytics. For simplicity, we treat all statistical methods as AI so the organization doesn’t have to worry about the specific methodology. With that definition, AI has been reshaping our industry for many years, but the pace is accelerating with generative and agentic AI. To me, this is the most fundamental change in the way we do business in 50 years. The biggest reshaping event in my business career is AI, and this holds true for everyone working today. In terms of competitiveness, the answer is clear: if you are not a frontrunner, you will be a laggard, and then you will be gone. We have no other choice but to invest heavily.
Q: Where are you seeing the biggest tangible impact from AI so far?
Eltze: Historically, like most retailers, we first saw tangible impact from analytical AI—particularly in marketing personalization and in forecasting and replenishment, improving both our offers and our supply chain. Today, the value is split across two perspectives. On the customer-facing side, AI continues to improve services, one-to-one personalization, and availability. Here, analytical AI remains the backbone, increasingly complemented by generative AI in areas such as content and service interactions. On the internal side, the biggest impact today comes from process efficiency. This is where generative AI and increasingly AI agents help simplify, speed up, and automate everyday work and end-to-end processes. My belief is that, in the long run, the customer-side impact will be higher, especially as agentic AI matures, because this side ultimately drives new revenue and margin, not just efficiency. However, realizing efficiency gains is still a huge, unproven opportunity for most large companies, and it will be a major focus for the next 12 to 24 months.
Q: What has been the hardest part of scaling AI across your organization, and how have you addressed it?
Eltze: There isn’t one single hurdle; the challenge is multidimensional. I see three core pillars we have to work on simultaneously. First is governance, which includes everything from security and compliance questions to clearly defined ownership and decision rights. These issues must be solved centrally. Second is technology selection. With the landscape moving so fast, we have to make a choice on a stable toolset to deploy across 380,000 people and then stick with it for a period to gain traction. The third, and perhaps most complex challenge, is the change management required to get hundreds of teams activated and motivated. Because this transformation is so fundamental, touching every process in every department, there is no central unit that can orchestrate it all. We have to empower each individual team—the marketing team here, the logistics team there, the store operations team in Hungary—to drive the change for themselves. Our role is to provide the tools and the framework, but the transformation must be decentralized.
Q: How is AI changing the skill requirements of your workforce, and how are you managing that transition?
Eltze: We are just at the beginning of this journey because large-scale reskilling is a direct consequence of realizing large-scale efficiency gains, which we have not yet fully achieved. So far, we have scaled up our central data science and AI teams to create a core group of experts who can navigate the organization through this change. Now, we are rolling out a broad communication and empowerment program. We recently put our top 200 senior leaders through mandatory, hands-on AI creativity workshops. The goal was not to turn them into technologists, but to help them understand what is possible, where the limits are, and how AI affects decision-making, processes, and leadership in their areas. My role is to orchestrate the overall approach, define the framework, and connect change management with the right toolsets—but I cannot do the transformation for them. We are now actively bringing these elements together to create real traction within the business units and to prepare the organization for the next phase of scaling AI impact.
Q: How do you think about risks around AI—data privacy, bias, or governance—and maintaining customer trust?
Eltze: From a customer trust perspective, we are very conscious of the risks around AI and treat them as an ongoing leadership responsibility. Today, we deliberately limit customer-facing AI use cases to areas where the risk of eroding trust is manageable and well understood. For example, when we use gen AI for product descriptions or hotel reviews, we have clear safeguards in place to prevent major errors or misleading outputs. For critical information like allergens, we always consistently ensure a human-in-the-loop approach to validate and cross-check the output.
At the same time, I see a very immediate and often underestimated internal risk. Tools like MS Copilot allow employees to generate polished outputs extremely quickly, and the danger lies in accepting them uncritically. AI can produce a convincing report in minutes, but without sufficient AI literacy and domain expertise, inaccuracies, bias, or gaps may easily go unnoticed. For me, this internal risk is closely linked to customer trust. If we do not educate our people properly and reinforce critical thinking and ownership, internal misuse or overreliance on AI will eventually show up in external outcomes. That is why strengthening governance, skills, and leadership behavior around AI adoption is just as important as managing customer-facing risks.
Q: How do you think AI will impact the competitive landscape and what will differentiate the winners over the next 3–5 years?
Eltze: I believe there are three key areas. First, we must continue to make a big leap on AI—getting better offers, better assortment, and deeper personalization to our customers, because that generates real, bottom-line value. Second is the once-in-a-lifetime opportunity to improve our internal efficiency and processes. Third, we have to monitor the rise of agentic e-commerce. For our tourism business, where a significant part of journeys may soon start on a platform like ChatGPT, it’s a serious challenge we need to address.
Ultimately, winning with AI is not just a question of money; it’s a question of culture, teams, and ways of working. Over the past few years, we have become a more data-driven company. That cultural asset is something competitors cannot simply buy. It puts us in a good starting position, but it is not a given. It will continue to be enormous work, but the good news is that unlike some other large investments, AI projects tend to have a fast and positive return.
Related Articles
The AI transformation manifesto
Rewired to outcompete
