In late 2021, IKEA’s parent company Ingka Group flipped a switch. An AI chatbot named Billie went live across IKEA’s customer service channels, handling everything from delivery tracking to return inquiries. Within months, Billie was resolving 47% of all incoming customer questions without a human ever getting involved. That freed up 8,500 call center workers across 22 countries who suddenly had a lot less to do.
Most companies would have handed them severance checks. IKEA retrained every single one of them as interior design consultants instead. Those former phone reps now sit on video calls with customers, walking them through 3D room renderings and personalized floor plans. The result: a design consulting service that generated 1.3 billion euros in revenue in its first full year and zero layoffs.
IKEA’s AI reskilling program is the largest documented case of a company using artificial intelligence to create a new revenue stream by retraining displaced workers rather than eliminating their jobs. The story went viral on LinkedIn in April 2026, racking up nearly 500,000 views in 24 hours, because it contradicts the dominant narrative that AI means fewer jobs. It doesn’t have to. But copying what IKEA did requires understanding what they actually did, and what most companies get wrong.
Last updated: April 2026
How did IKEA use AI without laying off workers?
IKEA didn’t start with a reskilling plan. It started with a chatbot rollout that accidentally revealed a billion-dollar opportunity hiding in its customer service data.
When Billie went live in FY2021, the chatbot handled level-one inquiries: order status, delivery times, return policies, basic product questions. It resolved 47% of all incoming queries without human escalation, saving an estimated 13 million euros across 3.2 million interactions between 2021 and 2023. By 2026, that resolution rate climbed to 57%.
But the real insight came from the 53% of inquiries Billie couldn’t resolve. When Ingka Group’s data team examined the patterns in those unresolved conversations, a signal jumped out. Customers weren’t calling about broken shelves or late deliveries. They were asking for help planning their living rooms. They wanted someone to tell them which KALLAX unit fits in a 12-foot wall. They wanted advice on paint colors that match their EKTORP sofa.
IKEA had thousands of customer service workers who already knew the product catalog inside out. The gap wasn’t product knowledge. It was design training. So Ingka Group built a reskilling program focused on three areas: remote interior design competency, digital retail sales, and complex customer relationship management. The 8,500 workers didn’t just get a new title. They got trained on 3D rendering software, video consultation techniques, and space planning methodology.
Parag Parekh, Ingka Group’s chief digital officer, described the approach as using AI to enhance rather than replace human capability. Ulrika Biesert, the company’s people and culture manager, framed it as a commitment to “lifelong learning and reskilling, and to accelerate the creation of new jobs.”
What is IKEA’s AI chatbot Billie?
Billie is IKEA’s AI-powered customer service chatbot, built on natural language processing technology and deployed across IKEA’s global customer service network. It operates 24/7 across multiple time zones and languages.
The chatbot handles a specific slice of customer interactions: order tracking, product availability checks, return and exchange procedures, delivery scheduling, and basic product information. These are high-volume, low-complexity interactions that follow predictable patterns. A customer asks “where is my order?” and Billie pulls the tracking data and responds in seconds.
What Billie can’t do matters as much as what it can. It doesn’t handle emotional purchasing decisions, complex design questions, complaints that require empathy, or any interaction where the customer needs to feel heard by another person. That division of labor is intentional. IKEA designed Billie to absorb the repetitive work that burned out human agents, not to replace the interactions where human connection creates value.
The name itself is a nod to IKEA’s product naming tradition (you’ll also see it spelled “Billy” in some coverage, after the company’s bestselling bookcase). The chatbot appears on IKEA’s customer service pages as the first point of contact, routing more complex issues to human specialists when it detects a conversation is beyond its scope.
The $1.4 billion design business nobody planned
The remote interior design service generated 1.3 billion euros in sales by the end of FY2022. That’s roughly $1.4 billion, and it represented 3.3% of Ingka Group’s total revenue in its first full year of operation.
Here’s how the service works. A customer books a consultation through IKEA’s website or app. They get matched with a design consultant (one of the 8,500 reskilled former call center workers) for a 1-on-1 video call. During the session, the consultant reviews the customer’s space, discusses their needs and budget, then produces a 3D rendering of the proposed design, a detailed floor plan, and a shopping list of IKEA products to fill it.
The average design consultation leads to a much larger basket than a standard IKEA shopping trip. When a customer walks into a store, they might buy a bookcase. When they sit through a design consultation, they buy the bookcase, the desk, the shelving, the lighting, and the rug. The human relationship and personalized advice drive higher-ticket purchases that a chatbot could never produce.
Ingka Group has set a target of growing this channel to 10% of total revenue by 2028. If the company’s revenue holds near its current roughly 40 billion euro range, that target means the design consulting business alone could hit 4 billion euros within two years. That’s a business line that didn’t exist before the chatbot launched.
The financial logic is straightforward. IKEA spent money on reskilling (training programs, software licenses, infrastructure for video consultations) but avoided severance costs for 8,500 workers, retained institutional knowledge, and built a service that customers are willing to pay premium prices for. The reskilling investment paid for itself within the first year of operation.
Why most companies take the opposite approach
IKEA’s story stands out because almost nobody else is doing this. A January 2026 Harvard Business Review analysis found that companies are laying off workers “because of AI’s potential, not its performance.” The cuts are preemptive, not evidence-based. The pressure on founders to pick a side is real, as the debate around AI leadership and transparency makes clear.
Consider the contrast with Medvi, the telehealth startup that made headlines in April 2026 for hitting $1.8 billion in projected annual revenue with just two employees. Matthew Gallagher built the entire operation using ChatGPT, Claude, and Grok for everything from code to customer service. He outsourced doctors, pharmacies, and compliance to third-party platforms. It’s the polar opposite of the IKEA model: maximize AI, minimize humans.
Both approaches generated massive revenue. But the tradeoffs are different. Medvi now faces growing scrutiny and negative allegations about quality and oversight. IKEA’s design consultants, meanwhile, are building long-term customer relationships that increase lifetime value and repeat purchases.
Then there’s the Burger King approach: using AI not to replace or reskill workers, but to monitor whether employees say “please” and “thank you” through a headset chatbot called Patty. That’s AI as surveillance, not strategy. It treats workers as problems to be policed rather than assets to be developed.
| AI approach | Example | Workers affected | Revenue impact | Risk |
|---|---|---|---|---|
| Reskill and redeploy | IKEA (Billie chatbot) | 8,500 retrained | +$1.4B new revenue | Training costs, slower rollout |
| Full AI replacement | Medvi (2-person startup) | Near-zero human workforce | $1.8B projected 2026 | Quality concerns, reputational risk |
| AI as surveillance | Burger King (Patty chatbot) | Monitored, not reskilled | Unknown | Employee morale, turnover |
What founders can steal from the IKEA playbook
IKEA is a 80-year-old multinational with 177,000 employees. You probably don’t have that. But the underlying logic of their approach scales down to a 5-person startup, and it comes down to three decisions.
First, map AI to your most repetitive tasks before touching anything else. IKEA didn’t try to automate design consultations or product development. It automated the interactions that followed scripts: order tracking, return policies, delivery updates. Identify the tasks in your business where the answer is the same 90% of the time. That’s where AI goes first. For a founder running a micro SaaS business, that might mean putting a chatbot on tier-one support tickets while you personally handle onboarding calls.
Second, look at what AI can’t resolve. That’s where the money is. IKEA’s breakthrough wasn’t the chatbot. It was the pattern in the chatbot’s failures. The 53% of unresolved queries told Ingka Group that customers wanted design help, not faster answers to shipping questions. When you deploy AI in your business, pay close attention to the tasks it struggles with. Those struggles are signals pointing to unmet demand. If you’re building products with AI-powered tools, the bugs your AI can’t fix are probably the problems your customers will pay you to solve.
Third, reskill before you cut. The default instinct when AI handles a task is to eliminate the person who used to do it. IKEA’s results show that’s leaving money on the table. Those 8,500 workers already had years of product knowledge and customer interaction experience. Replacing that institutional knowledge from scratch would have cost more than the reskilling program. If someone on your team is doing work that AI can now handle, ask what higher-value work their existing knowledge qualifies them for before you write the termination letter.
This isn’t idealism. It’s math. IKEA spent less on reskilling than it would have on severance, recruiting, and training new design consultants from scratch. And the reskilled workers generated $1.4 billion in revenue that didn’t exist before the chatbot launched. The creator economy runs on a similar principle: the people who thrive aren’t the ones replacing themselves with AI, but the ones using AI to move up the value chain.
Can other industries copy the IKEA reskilling model?
The honest answer: not directly. IKEA’s model worked because retail has a built-in upsell path from customer service to design consultation. A furniture company can turn a support agent into a design advisor because the jump makes intuitive sense. Not every industry has that path.
But the principle transfers. A law firm can automate document review with AI and reskill paralegals into client relationship managers. An insurance company can automate claims intake and retrain adjusters as risk consultants. A marketing agency can automate reporting and move analysts into strategy roles. The pattern is the same: AI absorbs the routine floor, humans move to the judgment ceiling.
The industries best positioned to copy this model share three characteristics: high customer interaction volume, a gap between current service and premium service, and workers whose existing knowledge gives them a head start on the premium role. If your business has all three, you’re sitting on the same opportunity IKEA found.
IKEA plans to train 70,000 employees in AI literacy by the end of 2026. That’s not just about the chatbot. It’s about building a workforce that understands where AI fits and where humans are irreplaceable. For founders watching the AI race from the outside, the lesson isn’t to hire fewer people. It’s to hire people who know what AI can’t do, and put them there. Companies like those building new communication platforms and those developing the AI models themselves are all grappling with the same question: where does the human add value that the machine cannot?
