The Maze: Walmart is turning Scintilla from a supplier data library into a decision engine. New and planned AI features will flag changes, answer business questions and guide action across inventory, media and store execution. Next year, brands selling to Walmart and through its Marketplace should also see both models in one view. The promise is speed: less time hunting through reports, more time acting on the signal. The strategic shift is larger. Walmart increasingly controls the data, the analysis layer and the route from insight to action.
The next product layer is explanation, not simply more reporting. At Inspire 2026, Walmart Data Ventures outlined personalized experiences, custom alerts and a more capable Marty assistant grounded in Walmart first-party data. The assistant is meant to surface performance drivers, explore complex questions and create dashboards around metrics a supplier wants to monitor. That sits on top of Scintilla Boards, launched in July to bring customer, sales and operational signals into a common view. The timing matters: Boards exist now, while several conversational and alerting features are still upcoming. Walmart is describing a product direction, not a fully shipped autonomous analyst.
Walmart is connecting diagnosis to systems where suppliers spend money or move stock. Its example is deliberately operational: Scintilla spots an item at risk of going out of stock in 50 stores, then sends the user toward the replenishment system. Scintilla Insights Activation already links findings with Walmart Connect, turning shopper and performance signals into media campaigns. The current product architecture also covers digital discovery, basket behavior, report building and customer research. The value comes from joining these surfaces. A signal about search, conversion or availability becomes more valuable when the same ecosystem offers the next action.
The store is becoming another endpoint for the intelligence loop. Scintilla In-Store launched in February as a mobile workspace for third-party supplier representatives, combining store data, item information, tasks and certifications. Walmart is now planning AI recommendations that tell those teams what needs attention during a visit. That could move the platform from observing execution to prioritizing it. But the existing app already handled fieldwork; September's announcement adds an intelligence layer rather than launching the operating surface itself. Suppliers should separate new capability from new packaging.
A combined first-party and Marketplace view changes the bargaining position. Brands that operate both as Walmart suppliers and Marketplace sellers currently manage different commercial models and data sets. The planned 2027 integration should show their Walmart.com assortment in one place. That can reduce blind spots and make cross-model decisions easier. It also deepens dependence on Walmart's definitions, permissions and recommended actions. The platform can tell a supplier what is changing, explain why and increasingly route the response. Walmart did not disclose pricing, model accuracy, feature-level availability dates or measured productivity gains. Those missing details decide whether intelligence becomes operating leverage or another subscription layer.
Why it matters: Retail data is abundant; coordinated action is scarce. Walmart is using proprietary customer and operational signals to close that gap inside its own ecosystem. Suppliers may get faster answers and fewer handoffs between analysts, agencies and field teams. Walmart gets a stickier data business and more influence over replenishment, retail media and store execution. Operators should judge the rollout by three things: which features are live, whether recommendations improve outcomes, and how easily decisions can be challenged or moved outside Walmart's stack.
Sources: PYMNTS | Walmart announcement | Walmart Data Ventures | Scintilla In-Store
