AI took over nearly half of IKEA's customer service work. But instead of laying people off, they reskilled them into new roles where they generated >€1.3B in a single year. When IKEA's "Billie" chatbot started resolving about 47% of customer inquiries - 3.2 million interactions, nearly €13 million saved - 8,500 call center co-workers suddenly had capacity freed up. That's the moment the corporate script usually writes itself - use AI to automate the work, cut the people, book the savings. IKEA did something more thoughtful and more difficult. They reskilled all 8,500 people with "remote interior design competence, digital retail sales, building relationships, and handling unique customer inquiries that require complex problem-solving," enabling them to perform work that work that needs taste, judgment, and real conversation. The remote selling business went on to generate €1.3 billion in a single year. The chatbot didn't reveal which jobs to eliminate. It revealed which human capabilities were being wasted on routine work. The AI questions aren't always "what can AI replace?" or "how can we use AI to save money," but "what are our people too buried to do?" and "how can we use our people to better serve our customers?" Most of the AI-and-jobs conversation treats automation as a subtraction problem. IKEA treated it as a reallocation and new business model exploration. One mindset shrinks an organization. The other grows it. If AI freed up 20% of your team's time tomorrow, do you actually know what higher-value work you'd point them toward? Or would the savings just quietly become the plan? I think this case study from Ikea is incredible and companies need to reflect on it. And btw - they did this over 3 years ago (in case you missed it like I did)!
Could I recommend a Netflix series? The title is "Legends", and it is based on the true story of low-level civil servants recruited to take down two massive drug cartels. The "recruiting process" identified the civil servants who could do more, e.g., search incoming suitcases through 3 weeks of training. It is a brilliant example of what upskilling can do with dormant talent!
I can feel a bit that populist energy myself, Glenn. Workers aren’t serfs or inventory; they’re free Americans who took the job knowing it’s at-will. The dignity of work cuts both ways: companies must steward resources wisely and employees build skills that survive downturns. Blame the planners? Absolutely. But a vast majority of these high-tech employees are compensated at levels that's inconceivable to most Americans. I hire about 8,000 a year, and there are executives in there, not many, my bet is our average salary annualized is about $45k. So while I empathize with anyone laid off, if they were earning like $300k at Facebook, working from home, then get a massive severance .... yeah they're doing better than 99% of the planet.
John D.
3d
1. The Cost and Difficulty of Redundancy: Under Sweden's Employment Protection Act (LAS), an employer cannot just dismiss people because an AI tool is more efficient. Redundancy (arbetsbrist) requires a lengthy legal process. Firing 8,500 people would require massive severance packages, months of paid notice periods (up to 6 months per person), and a massive legal burden. 2. The "Last In, First Out" Rule: Swedish law dictates strict seniority rules (turordningsregler). If IKEA wanted to downsize its call centers, they would have been legally forced to lay off their newest hires first, regardless of whether those individuals were the ones displaced by the chatbot.Mandatory Union Co-Determination: 3. Under the Co-Determination Act (MBL), IKEA was legally required to negotiate every step of this technological rollout with Swedish trade unions. Unions push heavily for internal redeployment and training over job cuts.
Donna Svei
3d
I'm working on a resume where a single AI implementation will easily generate $100+ million bottom line growth this year. Similar story, some people have new roles, but no one has been laid off. In this case, AI is doing something new that previous people + systems couldn't do.
Thanks for sharing this fantastic example of a people first culture Glen Cathey
I think AI as an industry itself will benefit more when they'll start talking about ROI use cases vs. doom and gloom use cases.
This is the case study most of the layoff-and-rehire crowd should be required to read. IKEA found the question almost nobody is asking right now, which is what are our people too buried to do, instead of what can AI replace. The piece that compounds is the second order effect you didn't even need to mention. The 8,500 people who got reskilled instead of cut are now the people who know the customer, the legacy systems, the edge cases, AND the new AI workflows. That stacked capability is what builds a moat. The companies running the subtraction play are stripping that exact layer out to fund infrastructure. One mindset shrinks an organization, the other grows it. That's the whole argument in one sentence and I'm stealing it.
I know quite a few people in the US that have been affected by Ikea's reorganizations over the past five years. I expect to hear that there will be additional layoffs later in the year.
Here’s the link to their official release: https://www.ingka.com/newsroom/ai-and-remote-selling-bring-ikea-design-expertise-to-the-many/