In September, Swedish shop-fitting group ITAB wrote off SEK 244 million (about €22 million) to stop its new ERP rollout — and decided to keep running the business on the systems it already owned. Three years into an IFS implementation that began under its "One ITAB" program, the board killed it and reverted to a combination of existing ERPs and group-wide applications. Wall Street would call that a failed transformation. I'd call it the most clear-eyed decision of the quarter.
The assumption worth retiring
For two decades, "digital transformation" has been shorthand for one move: rip out the core system and put in a modern one. Replace the ERP, the logic goes, and the rest follows. 2026 is finally burying that logic.
The receipts
ITAB is not an outlier; it's the pattern. Korea's KT halted its 1.6 trillion won (roughly $1.2 billion) KAIROS-X system overhaul barely a year in. Tennant's North American ERP go-live in February broke order processing and shipping, costing roughly $30 million in lost sales, a 23% stock drop, and now a federal securities investigation. Gartner warns that more than 70% of 2026 mainframe-exit projects will miss their targets because organizations overestimate what GenAI migration tools can do.
The real skill
None of these boards failed at transformation. They failed at replacement — and, in ITAB's case, had the courage to admit it. A core system encodes decades of order-to-cash, pricing, and compliance logic that a greenfield build can't reproduce on schedule. The discipline worth paying for in 2026 isn't another go-live playbook; it's knowing what not to replace.
Before you fund the next ERP business case, ask two questions: what does the current system do that the business actually depends on, and is the problem the system — or the process wrapped around it? Sometimes the most transformational line item on the P&L is the write-off.