IKEA Reskilled 8,500 Call Centre Staff. The Record Still Has to Answer.

IKEA did not add billions in extra sales by teaching people product data. Ingka, the largest IKEA retailer, reskilled 8,500 call centre co-workers after a chatbot took the routine questions, and its remote design channel turned over €1.3 billion in FY22. That figure is channel sales, 3.3% of the total, not a measured return on a PIM programme. The part a product-data lead can copy is narrower, and harder.

The primary account is Ingka's own note on AI and remote selling, with Reuters reporting the same figures on 13 June 2023. Billie, the chatbot named after the Billy bookcase, launched in FY21. From 2021 to 2023 it resolved about 47% of the customer enquiries it received: 3.2 million interactions, and nearly €13 million in savings. With the simple questions off the queue, Ingka retrained 8,500 call centre co-workers for remote interior design, digital retail sales, building relationships, and the enquiries that need a person. Ulrika Biesert, global people and culture manager at Ingka, told Reuters: "We're committed to strengthening co-workers' employability in Ingka, through lifelong learning and development and reskilling, and to accelerate the creation of new jobs." Sales through remote customer meeting points reached €1.3 billion at the end of FY22. Ingka's aim was to grow that share toward 10%.

A widely shared commentary, How IKEA Reskilled 8,500 Employees To Boost Sales by $1.4 Billion, reads the same remote-channel turnover as additional revenue caused by the reskilling. Ingka does not. Gross sales of a channel are not incremental profit, and they are not evidence that a product-information project paid for itself. Use the Ingka figures. Drop the causal leap.

What Ingka describes is a split in the work. A bot answers the question that has one answer. A person takes the question that needs judgment: which combination fits this room, this measurement, this previous order. That person is only as good as the record they can open while the customer is still on the video call. Dimensions, materials, compatibility, sustainability claims, what is actually available to promise. If those live in three systems and a private spreadsheet, the training produces a polite conversation and a wrong recommendation. The customer feels the record, not the org chart.

Ingka does not publish a PIM, DAM or ERP architecture for this programme. The training topics they do name (remote interior design, digital retail sales, relationships, complex enquiries) are customer skills. Product fluency sits underneath them, unnamed. That is the gap worth funding.

Three systems have different jobs, and the adviser needs all three without knowing their names:

• The product record holds the specs, attributes, translations and the claims you are willing to defend. What a PIM holds, and what a DAM holds, is the split between that record and the images, assembly guides and video the adviser shows on the call.
• Availability and cost sit next door. Where ERP stops and PIM starts is the argument for not making the adviser guess what is in the bay.
• The experience on the call is only as reliable as the record under it. What PXM stands for makes that order explicit: presentation after the record, not instead of it.

Product data quality and governance is where this gets operational. If the reskilled person repeats a wrong width, the remote channel scales the error to every customer who asked that week. The questions that keep coming back are your missing attributes. Write them into the record. Do not leave them in the head of the person who happens to be good at the call.

The trade-off is scale. Retraining 8,500 people is an Ingka programme, with a bot that already absorbs almost half the queue and a remote channel that already exists. A merchant team of 40 cannot copy the headline. It can copy the split on the ten questions customers actually ask, add the attributes those questions expose, and put that record in the hands of whoever still talks to the customer. That costs less than a new channel. It is also the part the €1.3 billion figure does not prove.