Key Takeaways:
- Invest in people before expecting adoption to scale. The organization realized that employees would not continue developing AI capabilities independently without intentional support. Leaders responded by investing in foundational training, AI champions, self-service learning resources, and recognition programs designed to build confidence and encourage experimentation.
- Create visible examples of success to make AI feel practical and accessible. Momentum increased when employees could see what peers were building and learn from those examples. Communities of practice, internal showcases, and recognition programs helped normalize experimentation while reducing the intimidation factor often associated with new technologies. Leaders also acknowledged that sustaining engagement after the initial excitement wears off remains an ongoing challenge.
- Use early projects to build organizational learning, not just business value. Rather than waiting for a perfect roadmap, the company used early AI deployments to better understand technical requirements, governance implications, licensing needs, and skill gaps. Those projects generated knowledge that informed future initiatives and accelerated organizational learning.
- Rethink processes before automating them. Teams were encouraged to start with a blank page rather than asking how AI could fit into existing workflows. The goal was to redesign work around desired outcomes and then determine where AI could add value. This approach helped avoid using new technology to simply accelerate outdated processes.
June 11, 2026: Highlights from the Supply Chain Board and Manufacturing Board’s member discussion, “How It Got Made: An AI Adoption Journey”
The Story in Brief
“One of the strongest lessons from this global fashion retailer’s journey was that AI momentum doesn’t sustain itself. Employees need support, examples, and opportunities to learn from one another. The organizations making progress are increasingly treating AI adoption as a change-management challenge, building communities and operating models that help experimentation become part of everyday work.” — Cheryl Goldsby, Membership Director, Supply Chain Board
Picking Up the Story
Building an AI Adoption Engine
Many organizations have spent the last two years experimenting with AI. The hard part is moving from isolated pilots to widespread, day-to-day adoption.
That was the challenge one global fashion retailer found itself facing. After early investments in AI-enabled business initiatives and employee productivity tools, the company realized it had assumed employees would continue developing AI capabilities on their own. Instead, leaders found that curiosity was not enough to sustain momentum.
The company’s response was to build a structured operating model centered on clear ownership, practical governance, training, AI champions, and peer learning. The discussion offered a behind-the-scenes look at how one organization is working to move from isolated AI projects to organization-wide adoption.
”"We realized we needed a strategy, because we had been treating AI as a simple use case and that people would be curious to continue learning themselves."
Focus of the Discussion
Scaling AI Required a Different Operating Model
The retailer’s AI journey began with early investments in data and AI capabilities, broad deployment of enterprise AI tools, and experimentation with emerging AI applications. As adoption expanded, leaders found that treating every AI initiative the same created confusion around ownership, governance, and execution. In response, the company developed a framework that separated everyday AI usage, technology-supported initiatives, and more advanced externally supported work. The structure provided a clearer way to evaluate risk, allocate resources, and determine where different types of AI work belonged.
As the organization moved into more advanced use cases, AI initiatives began exposing broader operational challenges. Data quality issues, fragmented knowledge, and information-management gaps often proved more limiting than the technology itself. Rather than designing a comprehensive roadmap upfront, the company chose to move quickly, learn from early projects, and refine its approach over time.
