top of page

Walmart Is Rethinking Advertising for Agentic Shopping

9 minutes ago
7 min read

Walmart is beginning to rethink how advertising works as AI shopping agents become capable of researching products, comparing options, and completing purchases on behalf of consumers. The company is not abandoning its fast-growing advertising business; instead, it is testing how brands can remain visible and useful when shopping shifts from human-led browsing to agent-led conversations and transactions. Walmart’s leadership has said it is still learning how conversational prompts differ from traditional search, but it sees an opportunity to develop advertising formats and discovery methods designed for agents.


Walmart Is Rethinking Advertising for Agentic Shopping

What agentic shopping changes

Agentic shopping refers to the use of AI systems that can act with some degree of autonomy on a shopper’s behalf. Rather than manually searching a website, comparing listings, and checking out, a consumer may ask an agent to find a product within a budget, evaluate alternatives, track prices, reorder household essentials, or complete a purchase. Consumer-facing agents such as Meta’s Muse can seek deals, compare products, and transact for users, which changes the role that conventional ads have traditionally played in guiding shoppers through a purchase journey.

This matters for retailers because much of modern retail media depends on a shopper actively searching, browsing, clicking, and viewing sponsored placements. If an agent handles those steps, the familiar sponsored-search ad or display banner may no longer be the primary way a brand reaches a customer. Walmart has acknowledged this shift directly: its executives have said that conversational commerce and the prompts people use in AI agents differ from the search behavior Walmart has historically understood.

The challenge is not simply that agents may reduce ad impressions. It is that agents may prioritize structured product information, availability, price, reviews, delivery options, and consumer instructions over visually designed marketing. Walmart has noted that agents may be less responsive to images or visuals intended to create an emotional response. That does not mean brand marketing disappears, but it suggests that brands will need new ways to make their products discoverable, understandable, and trustworthy to both shoppers and the AI systems acting for them.


Walmart’s advertising position

Walmart enters this transition from a position of commercial strength. In the three months ended October 31, 2026, its U.S. advertising revenue grew 37% year over year, supported by e-commerce momentum and assets including Vizio and Vibe.co. Walmart Connect, the company’s main U.S. advertising business, grew 36% in the quarter when excluding the effects of Vizio and Vibe.co. Globally, Walmart’s advertising business grew 25% in the same quarter, ahead of its 17% e-commerce sales growth.

Those results give Walmart a reason to be cautious rather than panicked. Its advertising business is still expanding, and its connected-TV assets give it reach beyond on-site sponsored search and display formats. Vizio’s TV unit volume rose nearly 150% year over year, according to Walmart’s chief growth officer, Seth Dallaire, expanding the audience available to advertisers through the platform’s operating system. Walmart has also integrated Vibe.co, a self-service TV advertising platform acquired in August, into its broader connected-TV offering.

But agentic shopping could eventually change where advertising dollars are spent. If shoppers increasingly rely on agents to narrow choices and make purchases, retailers may need to compete for visibility inside agent interfaces rather than only on their own websites. That raises difficult questions: How should sponsored recommendations appear in a chat thread? How will an agent decide which products to recommend? And how can advertisers measure whether an ad influenced a purchase that was completed by an AI system?

Walmart has not yet provided a complete answer. It has said it is considering how to enable brand advertising in these new environments, but it has not publicly detailed a final advertising model for agentic commerce, including how placements, campaign measurement, or attribution would work.


Building for agent-led commerce

Walmart’s response has been to build infrastructure that makes agent shopping possible while preserving customer control. In October 2026, it introduced Agent Profiles, the Walmart Agent Gateway, and terms for AI agents. These tools are intended to let customers shop with Walmart and Sam’s Club through external agents while setting permissions, spending limits, and approval requirements.

Agent Profiles give each authorized AI agent a dedicated profile within a customer’s Walmart account. Customers can see an agent’s purchases separately from their own, choose whether every purchase requires approval, or allow purchases up to a spending limit such as $50, $100, or $200. Walmart compares this to adding an authorized cardholder to a credit card: the agent can act within defined boundaries, while the customer retains the ability to monitor and revoke access.

The Walmart Agent Gateway is designed to give agents a more reliable way to interact with Walmart than navigating a website built for human users. Walmart says it will use an MCP, or model text protocol, beginning with a beta for companies and developers. The gateway is also intended to help agents find a customer’s usual items without exposing the customer’s broader personal information or purchase history. For example, an agent could help replenish a preferred brand of milk or sourdough bread without receiving the customer’s full shopping record.

Walmart’s stated terms include several protections. Customers must approve agent purchases by default, and Walmart may require additional approvals even for purchases within an agent’s spending limit. Agents cannot receive customers’ payment-card information or add or change payment methods. They can access only the information needed for a permitted task, and agents that violate Walmart’s terms can be blocked from its site. Walmart also says prices shown to agents are the same as those shown to human shoppers on Walmart.com, with applicable offers and fees displayed before purchase.

This infrastructure is commercially important because it creates a controlled pathway through which agents can shop at Walmart. It may also give Walmart a foundation for future advertising products, although the company has not said how or whether the gateway will affect ad placements or attribution.


Advertising inside Sparky

Walmart is also experimenting within its own conversational shopping assistant, Sparky. The company launched Sparky in June 2025, and usage in the United States grew 129% during 2026, according to Walmart. The assistant is designed to support customers from product discovery through purchase, using Walmart data and AI capabilities for functions such as item comparison, personalization, and shopping-journey completion.

Walmart began testing ads within Sparky in 2026. That makes Sparky an early laboratory for what agent-native retail media could look like: advertising embedded in a conversational experience rather than displayed beside search results. The company has not publicly laid out a detailed ad-product roadmap for Sparky, so it would be inaccurate to claim that Walmart has already built a mature agentic advertising marketplace. What is clear is that Walmart sees its own assistant as one place to learn how prompts, recommendations, and commercial messaging can coexist.

Walmart has also partnered with Google and OpenAI to enable discovery of Walmart products in their agentic shopping interfaces. At the Goldman Sachs Communacopia + Technology Conference in September 2026, Walmart described these partnerships as early and relatively small, but strategically important. The company’s objective is to ensure its assortment, prices, and shopping experience are available wherever customers choose to shop, including through outside AI platforms.


What brands may need

For advertisers, the immediate implication is that product data becomes more important. An agent needs accurate, structured information to determine whether a product matches a shopper’s needs. That includes price, availability, specifications, sizing, ingredients or materials where relevant, delivery timing, return policies, and reviews. Brands with incomplete, inconsistent, or poorly structured product information may be harder for agents to evaluate and recommend.

Walmart has already signaled that it is exploring agent-specific SEO alongside advertising strategies tailored for agents. This does not replace existing advertising; Walmart describes it as complementary. In practical terms, brands may need to optimize not only for human search terms but also for the questions, constraints, and preferences shoppers give agents. A shopper may not search for “best budget coffee maker”; they may tell an agent, “Find a highly rated coffee maker under $75 that ships quickly and has good reviews.” The agent then needs reliable data to interpret that request.

Brand advertising may still matter because consumers continue to want discovery and novelty. Dallaire has said that consumer curiosity about new products and brands has not disappeared even as AI changes how people explore options. The open question is how a brand earns attention when the shopper is not scrolling a page but delegating research to software. Sponsored recommendations, verified product attributes, rich product content, and measurable performance signals may become more central than conventional visual creative.


A cautious, early transition

Walmart’s approach is deliberately measured. Its executives have said that agentic shopping remains early and that volumes are not yet high. That caution is sensible: AI agents may reshape shopping eventually, but adoption, consumer trust, merchant participation, and advertising standards are still developing.

The most realistic reading is that Walmart is preparing for a future in which advertising becomes less about interrupting a human shopper’s browsing session and more about helping an AI agent make a useful, trustworthy recommendation. The company is building the technical rails for agent access, testing ads in its own assistant, partnering with external AI platforms, and studying how conversational queries differ from traditional search.

What remains unresolved is the business model. Walmart has not announced a definitive system for sponsored agent recommendations, nor has it explained how advertisers will be charged, ranked, or measured in agent-led journeys. Until then, the company’s strategy is best understood as experimentation and infrastructure-building rather than a completed reinvention of retail media.

Still, the direction is significant. If agentic shopping becomes mainstream, the most valuable advertising real estate may no longer be a search-results page or a product-listing grid. It may be the moment an AI agent decides which products are relevant, affordable, available, and trustworthy enough to present to a customer. Walmart is trying to make sure it has a role in that moment.


Executive summary

  • Walmart is preparing its advertising business for agentic shopping, where AI agents research, compare, and potentially purchase products on behalf of consumers.


  • The company has introduced Agent Profiles, the Walmart Agent Gateway, and AI-agent terms, while testing ads in its Sparky assistant and enabling product discovery through Google and OpenAI.


  • Walmart’s ad revenue remains strong up 37% year over year in the U.S. and 25% globally but it has not yet finalized how advertising, ranking, or measurement will work in agent-led shopping.


Sources

Comments


bottom of page