Levi's ERP Overhaul Lays Groundwork for AI Strategy, CDTO Says

What happened

Levi Strauss & Co's Chief Digital and Technology Officer, Jason Gowans, has set out the retailer's approach to modernising its core technology estate, centring on a global ERP overhaul and broader data modernisation programme. The initiative, discussed in an interview with Diginomica, is designed to give the denim maker a more consistent, connected data foundation that can support both employee-facing tools and an emerging AI strategy aimed at shoppers.

According to Gowans, the modernisation work is not simply an infrastructure refresh but a deliberate step toward enabling AI use cases across the business — from equipping staff with better systems and information to improving how the brand engages customers. The ERP and data programme is being framed as the groundwork that has to be in place before more advanced AI capabilities can be reliably deployed at scale.

Why it matters

This is fundamentally a digital transformation story about sequencing: Levi's is treating data and systems modernisation as the prerequisite for AI, rather than bolting AI onto legacy infrastructure. For large, established retailers with decades of accumulated systems and regional variation, a global ERP standardisation effort is often the unglamorous but necessary step that determines whether AI initiatives later succeed or stall on inconsistent, fragmented data.

For technology and transformation leaders, the signal here is that AI ambition is increasingly being used to justify — and prioritise — foundational data work that might otherwise be deprioritised as "just IT". Framing ERP modernisation as an AI enabler, for both internal teams and customer-facing experiences, gives such programmes clearer business sponsorship and a more compelling narrative for investment.

The Renascence take

The instinct in most organisations is to lead with the AI use case — a chatbot, a personalisation engine, a copilot — and treat the underlying data plumbing as a technical afterthought. Levi's approach inverts that order, and it's the more defensible one, even if it is less exciting to announce.

Most companies fail at AI not because the models are weak, but because the data feeding them is inconsistent, duplicated or locked in regional silos — and no one wants to fund the "boring" ERP project that would fix it. What's notable here is the discipline to sequence data modernisation ahead of AI, rather than retrofitting intelligence onto broken foundations. The behavioral lesson for leadership teams is that AI credibility is earned in the data layer long before it's felt in any customer-facing moment; operators serious about AI-driven experience should be prepared to sell unglamorous infrastructure work internally as the actual AI strategy, not a preamble to it.