The search bar, long the undisputed engine of digital commerce, is facing an existential evolution, and the world’s largest retailer is moving aggressively to monetize whatever comes next. Walmart has quietly begun exploring advertising formats within its new artificial intelligence shopping assistant, Sparky, initiating a strategic pivot that could redefine the economics of retail media networks. According to people familiar with the matter, the retail giant conducted a secretive test dubbed “Sponsored Prompts” beginning in September, a move designed to intercept high-intent shoppers as they migrate from keyword searches to conversational commerce. This initiative represents a critical defensive and offensive maneuver as Walmart seeks to protect its dominance against Amazon’s encroaching AI capabilities and the rising tide of agentic shopping.

The pilot program, which concluded earlier this month, allowed select advertisers to insert themselves into the dialogue between Walmart’s customers and its generative AI interface. Sparky, which was integrated into the Walmart mobile app in June, operates similarly to ChatGPT but is fine-tuned on Walmart’s vast product catalog and consumer data. It can summarize reviews, compare distinct items, and suggest purchases for specific events. However, the introduction of advertising into this ecosystem signals that Walmart views AI not merely as a customer service enhancement, but as the next frontier of high-margin revenue. As detailed in a sales presentation viewed by The Wall Street Journal, the new ad units are designed to leverage Sparky’s conversations for “brand engagement and personalized product recommendations,” marking a departure from the static display ads that currently populate e-commerce grids.

Converting Conversation into Commerce: The Mechanics of the ‘Sponsored Prompt’ and the Shift Toward Intent-Based Advertising

The mechanics of the test reveal how retailers plan to navigate the delicate balance between helpful AI advice and paid promotion. In the trial runs, users interacting with Walmart.com or the app were served a “Sponsored Prompt.” If a user clicked this prompt, Sparky would generate a contextual response followed immediately by a click-to-buy advertisement. For instance, a user might encounter a prompt asking, “What energy drink has the most caffeine?” Upon clicking, the AI would provide a factual answer regarding caffeine content, while simultaneously displaying a purchasable link for a specific brand’s energy drink. This format attempts to solve the “attribution problem” in AI chat, giving brands a direct line to consumers who are in the research phase of the buying journey but have not yet committed to a specific product.

This development comes at a pivotal moment for the retail industry, which is bracing for the widespread adoption of “agentic shopping.” Unlike traditional e-commerce, where users manually filter through lists, AI shopping agents act autonomously to curate choices and, eventually, execute transactions on behalf of the user. By inserting ads directly into the agent’s logic flow, Walmart is attempting to secure its share of trade dollars in a future where consumers may spend less time browsing catalog pages. eMarketer estimates that Americans will spend over $1.27 trillion on e-commerce transactions this year, a nearly 17% increase from 2024. As the volume of digital spend grows, the battleground is shifting toward controlling the interface that dictates those spending decisions.

The Amazon Factor: How Rufus is Forcing Walmart’s Hand in the High-Stakes Race for AI-Driven Revenue

Walmart’s urgency is partially driven by the rapid advancements of its primary rival, Amazon. The Seattle-based e-commerce titan launched its own AI assistant, Rufus, last year, and the financial implications of the tool are already staggering. During Amazon’s most recent earnings call, CEO Andrew Jassy revealed that Rufus is “on track to deliver over $10 billion in incremental annualized sales.” Named after the Welsh corgi that once roamed Amazon’s early warehouses, Rufus has already engaged with 250 million active customers. More critically for advertisers, Jassy noted that shoppers who utilized the chatbot were 60% more likely to complete a purchase. Amazon has already begun rolling out sponsored prompts to Rufus in North America, setting a precedent that Walmart is now compelled to follow to remain competitive for ad dollars.

The disparity in maturity between the two platforms highlights the immense stakes involved. While Amazon has successfully integrated Rufus into the purchase path for hundreds of millions of users, Walmart’s Sparky is still in its nascent stages. People familiar with the Walmart test indicated that the volume of users engaging with and clicking on the Sparky test ads was relatively low. However, this has not dampened advertiser appetite. Brands are desperate to understand how their products will surface in a world dominated by large language models (LLMs). Simon Poulton, executive vice president of innovation at digital marketing firm Tinuiti, noted that advertisers have been eager to test these formats to gauge how agentic shopping might alter sales velocity, even if current volumes remain modest.

The Economic Imperative: Why High-Margin Ad Revenue is Essential for Walmart’s Future Profitability

Beyond the technological arms race, Walmart’s push into AI advertising is underpinned by a fundamental economic imperative. The retailer’s core business—selling groceries and general merchandise—operates on notoriously thin margins. In contrast, its digital advertising arm, Walmart Connect, generates substantial profit margins that help subsidize low prices on store shelves. Walmart’s global ad revenue grew by a massive 53% in the most recent quarter, a figure bolstered by its acquisition of TV-maker Vizio. By turning Sparky into an advertising vehicle, Walmart is looking to open a new inventory stream that is immune to the physical constraints of store shelves or the layout limitations of a mobile app screen.

The integration of ads into AI agents also serves as a hedge against the potential disruption of traditional search advertising. For nearly two decades, retail media networks have built their fortunes on keyword bidding—brands paying to appear at the top of search results for terms like “laundry detergent” or “running shoes.” Generative AI threatens to upend this model by providing a single, synthesized answer rather than a list of options. If Sparky tells a customer that a specific detergent is the best value based on reviews, rival brands lose the opportunity to be seen unless they can buy their way into that conversation. The “Sponsored Prompt” is Walmart’s first attempt to create a commercial structure for this new reality.

Strategic Partnerships and the Broader AI Ecosystem: From OpenAI to the Store Floor

Sparky is merely the consumer-facing tip of a massive, company-wide artificial intelligence overhaul at Walmart. The retailer has engaged in a strategic partnership with OpenAI, a deal that is set to enable U.S. users of ChatGPT to purchase Walmart products directly within the OpenAI platform in the near future. This creates a bifurcated strategy: Walmart is building its own walled garden with Sparky while simultaneously ensuring its inventory is accessible through the world’s most popular external AI chatbots. This dual approach is designed to capture sales regardless of where the consumer begins their journey, whether it is inside the Walmart app or on a general-purpose AI platform.

Furthermore, the data powering Sparky is being leveraged across the enterprise. Walmart has deployed AI agents to assist associates, developers, and suppliers, creating a feedback loop of data that refines the shopping agent’s capabilities. By analyzing supply chain data and associate inputs, Sparky can theoretically offer more accurate stock information and fulfillment options than competitors. However, the success of monetizing this ecosystem rests on consumer trust. If shoppers perceive Sparky as a biased salesperson rather than a helpful assistant, engagement could plummet. The company has stated it is “considering next steps” following the conclusion of the September test, suggesting a cautious rollout as they analyze the data on user sentiment and ad relevance.

Navigating the Uncharted Waters of Agentic Commerce and the Future of Digital Retail Media

As the industry moves toward 2025, the distinction between a search engine and a shopping agent will continue to blur. The “Sponsored Prompt” test indicates that Walmart is unwilling to cede this territory to Amazon or emerging tech players like Perplexity or Google. The challenge for Walmart will be scaling these ad units without degrading the user experience. Unlike a search results page, which can accommodate multiple sponsored tiles, a chat interface has limited real estate; a conversation cluttered with ads quickly loses its utility. The low engagement numbers in the initial test suggest that finding the right context and timing for these prompts is a problem that has not yet been fully solved.

Ultimately, the monetization of Sparky represents a microcosm of the broader transformation sweeping the retail sector. As The Wall Street Journal reported, the goal is to leverage AI to combine detailed product info with customer needs, boosting online sales in an era where efficiency is paramount. Whether Sparky can generate the multi-billion dollar incremental sales figures seen by Amazon’s Rufus remains to be seen, but Walmart’s willingness to experiment with ad formats this early in the product’s lifecycle demonstrates a clear commitment to the strategy. For industry insiders, the message is unambiguous: the chat window is now open for business, and the race to own the AI-mediated checkout is officially underway.