Levi's ties AI strategy to global ERP and data modernisation

What happened

Levi Strauss & Co.'s Chief Digital and Technology Officer, Jason Gowans, has set out how a global enterprise resource planning (ERP) and data modernisation programme underpins the retailer's approach to artificial intelligence. Speaking about the initiative, Gowans described the overhaul as the structural foundation that will support AI applications for both employees and shoppers.

According to the account, Levi's is treating the consolidation of its core systems and data as a prerequisite rather than a side project — the view being that AI tools, whether used internally by staff or externally by customers, are only as reliable as the data and processes feeding them. The programme spans the retailer's global operations, suggesting a long-term, enterprise-wide commitment rather than a localised pilot.

Why it matters

This is fundamentally a digital transformation story: it illustrates how a legacy retailer is sequencing its modernisation, putting ERP and data cleanup ahead of visible AI deployment. That ordering matters for any organisation weighing whether to chase AI use cases immediately or first fix the plumbing that will determine whether those use cases actually work at scale.

For technology and experience leaders, the signal is that AI ambitions — whether aimed at improving employee productivity or shopper-facing services — are gated by the quality and unity of underlying data. A retailer the size of Levi's operating across multiple markets and channels cannot personalise, automate or forecast reliably if its systems of record are fragmented, so the ERP programme becomes the enabling layer for everything that follows.

The Renascence take

The instinct in many boardrooms is to lead with a flashy AI pilot and worry about data architecture later. Levi's approach — data and systems first, AI second — is a useful correction, and one worth naming plainly rather than treating as a footnote to the AI conversation.

Most organisations underestimate how much of "AI transformation" is actually unglamorous data and systems work — and overestimate how much value a chatbot or recommendation engine can create sitting on top of fragmented ERP data. The behavioral lesson is that trust, whether from an employee using a new tool or a customer receiving a personalised offer, breaks the moment the underlying data is wrong or inconsistent, and that trust is very hard to rebuild once lost. Operators serious about AI should audit their core data and process foundations before greenlighting customer-facing AI features, not after.