Swedish company IKEA fundamentally changed how the world’s consumers buy furniture. Now, the company is implementing agentic AI to fundamentally change how consumers interact with it, moving from transactional to immersive relationships. In this conversation with McKinsey’s Hai-Ly Nguyen and Holger Harreis, the chief digital officer at IKEA parent company Ingka Group, Parag Parekh, provides insight into the retailer’s journey and why prioritizing AI initiatives “is as important as the ideas themselves.” The interview has been edited for clarity and length.
The success is not the build. The success is deployment and adoption.
Q: How is AI reshaping IKEA’s strategy and competitive advantage today?
Parag Parekh: We look at the impact of AI across three distinct buckets. The first is customer experience, where both generative AI and agentic commerce are fundamentally changing how our customers interact with us, shifting from transactional relationships toward immersive, solution-oriented ones. The second is supply chain and logistics, where we are deploying AI to take costs out while staying close to the customer experience amid broader forces of localization, regionalization, and sustainability. The third is back-office productivity, where the goal is to free up our coworkers from repetitive tasks so they can redirect their time toward being customer-facing.
Q: How are you prioritizing across AI use cases given the scale of the opportunity?
Parekh: We use a four-quadrant framework that maps initiatives across two axes: customer-focused versus coworker-focused and growth-oriented versus cost-oriented. Use cases such as recommendation engines, pricing, and agentic commerce sit in the customer-growth quadrant, while supply chain and last-mile optimization sit in the cost-focused quadrant. We assess each initiative for its business and customer experience impact and deliberately stack-rank the portfolio from there. The discipline of prioritization is as important as the ideas themselves. The risk of doing everything but maybe nothing is very real at this stage, so we are very careful to concentrate and focus on the topics we want to go after rather than allowing AI initiatives to mushroom.
Q: Where are you seeing the biggest tangible ROI from AI today?
Parekh: The strongest returns are still coming from traditional AI use cases that have had several years to mature. The most concrete example is what we call goal-based order allocation: With more than 400 fulfillment and shipping points, we use AI with weightages to determine the optimal fulfillment route for each order, and that has meaningfully improved our cost ratio in shipping and logistics. Cross-sell and upsell personalization is another area of strong, compounding ROI, in which we use AI to serve customers relevant recommendations based on their context and shopping journey. On the generative AI side, customer service is emerging as a promising next frontier, particularly for multilingual, delanguaging capabilities and the automation of routine interactions, though we are not yet at a point of phenomenal scale there.
Q: What does the future customer experience look like at IKEA, and how does AI enable it?
Parekh: What we are seeing is a fundamental shift in customer intent. Historically, a customer might come to our website knowing they want a sofa or a dining table. But increasingly, the intent is changing. Customers are saying, “Help me design my living room. Help me shape my life at home.” And this is where AI combined with technologies such as LiDAR1 room scanning creates a completely different kind of interaction. We ask a few simple questions about style, color, and budget, and using the 3D scan of your room, we can generate four or five inspiring options to start from. You pick one, you refine it, you tell us what you do not like, and we iterate together. Suddenly, we have gone from a transaction-based relationship to helping solve a customer’s problem. And this will go further still into voice-based and chat-based agentic formats. That is how fundamentally AI will change the customer experience in this industry for us.
Q: What has been the hardest part of scaling AI across a 160,000-person organization?
Parekh: With any technology, one of the biggest challenges is always managing the change. A capability is only as good as its deployment and adoption, and that is what we measure end to end for every digital product. Some people see AI as an opportunity, some as a threat. So our goal is to help our co-workers better understand AI and what it offers, providing them with the tools to better perform their roles. We have started senior-leadership-onboarding workshops, so far completed AI literacy programs for 40,000 coworkers, and launched pilots in three countries where we work with store-level ambassadors to identify and showcase high-impact use cases.
Q: How is AI changing the skills and roles required across your workforce?
Parekh: I start with my own group digital organization as a leading indicator because this is where the change is most visible. The traditional software-development life cycle—from requirements through build, test, and deploy—is being compressed by AI into rapid prototyping and experimentation cycles. The role of what a product manager, experienced designer, or engineer looks like tomorrow is fundamentally changing as a result.
We are working through this under what we call Digital Next, asking what group digital should look like in 2028 and how we start preparing for that today. But the same imperative applies to every function. If you are in HR, finance, marketing, or sales, you need to ask: Which of my processes will AI fundamentally reset? What new opportunities will emerge that were previously too expensive or not possible? I do believe roles will evolve, and I believe they will evolve for good.
Q: How do you see the workforce impact of AI, and where is the opportunity versus the disruption?
Parekh: On the customer-facing side, AI is creating services that were previously impossible at the IKEA price point. Historically, helping a customer design a space took on average six hours of a coworker’s time and cost the customer around €70 per session. Through automation, that time has been reduced from six hours to approximately 30 minutes per room per coworker. On the back-office side, cost pressures over recent years have already led us to optimize and reduce, and the quality of services from certain functions has dropped as a result. If I can drive productivity through AI, I can free up a coworker or leader to actually sit down with me on topics where they add meaningful value rather than being consumed by repetitive tasks. It can allow us to offer far more quality services. And it will be something to keep a very close eye on as it evolves.
Q: What is your verdict on agentic commerce, and how should IKEA respond?
Parekh: It starts with a longer-term shift that has been under way for 15 to 20 years. There were days when brands decided how they met customers, but customers have been deciding where they meet brands for some time now, and brands have had to choose whether to show up or not. What we are seeing from the past two to three years is that the share of search has shifted drastically from traditional toward agentic search, which tells us customers are embracing the agentic experience. For brands, showing up in those experiences is likely to become necessary. The unresolved question, which mirrors the historical tension between traditional search and brand-owned assets, is whether commerce ultimately gets captured on the agentic platform or redirected back to the brand. That is the play that will unfold over the next two to five years. I would probably be naive to say exactly where it is going, but I believe there is interest on both sides to continue collaborating rather than polarizing one way or the other.
