Half of Walmart’s App Users Have Already Met Your New Buyer

On Walmart’s Q4 FY26 earnings call in February, CEO John Furner offered a figure that should sit at the top of every supplier’s planning agenda. Customers who use Sparky have an average order value roughly 35% higher than customers who don’t. Walmart U.S. CEO David Guggina added that about half of all Walmart app users have now interacted with the tool. That is not a beta metric from a limited test. That is a material channel producing measurable commercial outcomes at scale, nine months after launch.

Sparky synthesizes reviews, interprets natural language queries, and surfaces products it can construct a confident recommendation around. Guggina described it on the call as helping Walmart “evolve from traditional search to intent-driven commerce.” For suppliers, that phrase carries a specific operational meaning. Intent-driven commerce is not keyword commerce. It is not rank commerce. It is not advertising commerce in the traditional sense. It is a system that evaluates whether it can justify a recommendation before it makes one, and the content behind your listing is what that system reads when it decides.

Winning the Sparky Recommendation Is Not the Same as Winning the Auction

Walmart Connect in the U.S. grew 41% in Q4 FY26, part of a full-year global advertising total of $6.4 billion. Those numbers make it easy to assume that more Connect spend equals more Sparky visibility. That assumption will cost suppliers money.

Walmart Connect confirmed in January 2026 that it is testing sponsored prompts inside Sparky, meaning a brand can pay to appear in the conversational recommendation flow. Placement inside the conversation is not the same as being recommended. A sponsored prompt gets Sparky to consider your product. Whether Sparky actually recommends it depends on whether the system can assemble a coherent, defensible answer from your content. If the content is ambiguous, inconsistent, or built to communicate brand positioning rather than real-world outcome, Sparky can surface the sponsored prompt and still route the shopper elsewhere.

Furner framed the 35% AOV premium in terms of discovery quality, not paid placement. Sparky, he said on the call, is “helping customers find the things they need, they want, and they love” and is “strengthening our digital unit economics as it scales.” That is the language of confident, organic selection. The suppliers inside that premium are there because Sparky could build a case for their product, not because they outbid a competitor.

What Sparky Is Actually Doing When a Shopper Asks a Question

Sparky was designed to handle natural language queries, not keyword strings. When a shopper asks something like “what’s a good protein powder for recovery on a budget,” the system is not matching terms. It is interpreting the query as a set of constraints, price sensitivity, recovery-specific formulation, likely use frequency, and evaluating which products in Walmart’s catalog can be matched to those constraints with the least remaining uncertainty.

Products with incomplete attributes, vague use-case language, or descriptions that communicate aspiration rather than outcome create uncertainty. Uncertain products get filtered before the shopper sees any results. For 1P suppliers, that filtering happens against catalog data submitted through Supplier One. For 3P Marketplace sellers, it happens against content in Seller Center. The mechanism differs. The vulnerability is the same.

A feature list tells a system what a product has. It does not tell the system what problem the product solves, in what context, for what kind of shopper, with what trade-offs acknowledged. Sparky needs the latter to construct a recommendation it can stand behind. Lifestyle imagery serves human browsing well. It provides no structured signal for a system comparing products against a shopper’s stated constraints.

The Content Gap Most Suppliers Don’t Know They Have

Walmart’s Marketplace Learn documentation has long instructed sellers to provide content that helps shoppers feel confident they are making the best choice. That standard has not changed. What has changed is that a machine is now applying it before the shopper arrives.

The gap shows up most acutely in categories where shopper queries are inherently contextual: health and wellness, food and beverage, home improvement, pet care, consumer electronics. These are the categories where shoppers are most likely to describe a situation rather than name a product. A shopper asking “what’s the quietest blender for a small apartment” is giving Sparky a set of constraints to work with. Whether your blender surfaces in the answer depends on whether your content connects those specific attributes, noise level, footprint, kitchen context, to the product in explicit, outcome-linked language.

Suppliers who have not audited their PDPs against that standard are invisible to the selection logic driving the 35% order value premium.

High Review Volume Is Not the Same as High Recommendation Probability

Sparky synthesizes reviews as part of its evaluation, which has led some suppliers to treat review volume as an agentic visibility strategy. Reviews tell Sparky how previous customers experienced the product. They do not tell the system why the product fits this particular shopper’s situation. High volume cannot compensate for content that fails to connect product attributes to real-world use cases.

The more productive approach is ensuring that the language shoppers use in reviews, the specific situations, outcomes, and comparisons they describe, is also present in the structured content on the PDP. When a review says “perfect for small kitchens” and the product description also addresses footprint and noise level, Sparky has agreement across sources for the same claim. Agreement reduces uncertainty. Reduced uncertainty is what gets a product recommended.

Sparky Does Not Know Your Buyer Relationship

Sparky does not know your trade spend history, your years on shelf, or the strength of your relationship with your category merchant. It knows what your content says and whether that is enough to build a recommendation around. That evaluation runs the same way for a brand that has been on Walmart shelves for twenty years as it does for a Marketplace seller in their first quarter.

Walmart’s Q4 FY26 results establish that this channel is past the experimental stage. Half of app users, roughly 35% higher order values, 41% Connect growth, and a CEO who described Sparky on an earnings call as central to Walmart’s digital unit economics. The suppliers who treat content readiness as a 2027 problem are ceding ground in a channel that is already deciding whose products get found.

Winning With Walmart

Winning With Walmart is an independent platform for suppliers, sellers, solution providers, and industry experts.

Articles are developed by staff editors, contributing experts, and trusted partners.

Some are researched and drafted with the assistance of AI tools and are reviewed, fact-checked, and edited by Winning With Walmart editors before publication.

Editorial judgment, sourcing decisions, and final approval rest with the publication's human editors in every case.

We are committed to accuracy and fairness. If you believe this article contains an error, we welcome your feedback.