Walmart's Sparky: Agentic AI and the Future of Shopping

In June 2025, Walmart launched Sparky, its agentic AI shopping assistant designed to replace keyword search with a true service experience. Instead of browsing product lists, customers could state goals, such as hosting a cookout or planning weeknight dinners, and Sparky would plan, reason, and act to deliver a complete solution. Sparky could compress the consumer journey from intent to fulfillment by working across Walmart's retail stack (catalog, inventory, pricing, fulfillment, services). Sparky represented a strategic bet for Walmart as it moved to redefine shopping search journeys, strengthen customer loyalty, and unlock new monetization models through AI. The launch came at a critical juncture. Walmart faced intensifying competition from Amazon's Rufus and Instacart's Ask Instacart, both of which were embedding generative AI into their retail operations. Walmart's differentiation rested on two pillars: agentic depth--multi-step planning and action rather than single-shot answers--and openness, by designing Sparky to interoperate with other agents rather than locking customers into Walmart alone. The case places students in the shoes of Walmart's executive team as they consider three strategic dilemmas. Should Sparky prioritize customer experience or retail media monetization through advertising revenues? How far should Walmart go in personalization without making customers uncomfortable about privacy? Should Walmart build proprietary AI capabilities to maintain control over and differentiation in its technology stack, or should it partner for speed and focus?