Levi's began this project in 2019 with nine separate back-office management systems, each one customized by region, each one holding data in a different format. The company's CFO joined 13 years ago and recalls the same problem being discussed from day one. It took until now to actually fix it.
The platform they chose is SAP S/4 Fashion, running on Microsoft Azure's private cloud. Rather than bending the software to match old habits, Levi's adopted SAP's standard way of doing things. That decision, which required a full year of retraining staff to stop requesting custom features, is now the reason the company can move fast. SAP upgrades that typically take the industry 48 hours to complete take Levi's around 20 minutes, because there is no custom code creating complications.
The numbers tell the scale of the cleanup. More than 90 legacy systems have been retired. Over 2,600 employees now work from a single platform with a shared dataset. More than 80% of the company's business processes have been standardized globally. The remaining regions, Europe, Brazil, and the Andes, are expected to finish by mid-2027.
But the more immediate story is what this foundation now enables. AI software is already running inside the ERP, automatically processing sales orders, capturing invoices, and managing vendor compliance tasks that used to require a person. In parallel, Levi's has built specialized AI assistants across HR, finance, IT, and retail operations, and is now connecting them through a single interface inside Microsoft Teams. An employee checking inventory, filing an IT request, or pulling last year's budget data asks one system and gets one answer. The system routes the request to whichever underlying tool handles that task.
Store staff also have a separate mobile tool called Stitch, which gives them on-demand access to product information, training materials, and customer service guidance. A consumer-facing styling tool called Outfitting is live across the U.S., Canada, and major European markets.
For anyone running a business that relies on SAP's older software, the Levi's story carries a direct implication. SAP ends mainstream support for its legacy ERP system in December 2027, meaning no new security patches, no compliance updates, and no bug fixes after that date. Gartner estimates that at the end of 2024, only about 39% of companies on the older system had licensed the new one. At the current pace of migration, Gartner projects around 17,000 companies, nearly half the customer base, will still be running the old system past the deadline.
The problem is timing. A complex migration across multiple countries and systems takes anywhere from 18 to 42 months. Any company that has not started planning is already in a difficult position. As the 2027 deadline approaches, demand for qualified migration consultants will rise sharply, which means costs go up and timelines get harder to meet.
Levi's avoided that crunch by starting early and committing fully. The payoff is not just a cleaner back office; it is the ability to deploy new tools continuously and at speed. The lesson is straightforward: data infrastructure is what determines whether AI tools are useful or useless. Companies that run fragmented systems with inconsistent data will keep watching pilots fail. Companies that do the unglamorous work of cleaning up their foundations first are the ones that end up with automation that actually runs.