Getting Recommended By Sparky Is Now Something You Can Hire Out

At the Cannes Lions festival on June 23, Flywheel, the commerce agency owned by Omnicom, launched a service it calls generative engine optimization, built to get brands recommended inside the AI shopping assistants run by Walmart, Amazon, and Target. The capability audits and rewrites product-page content for the signals those assistants weigh, and Walmart’s Sparky is named alongside Amazon’s Alexa for Shopping and Target’s assistant. Omnicom described the launch as the newest addition to a portfolio of generative engine optimization services it now runs across commerce, media, and public relations.

The product detail matters less than the signal it sends. Optimizing for an AI assistant has moved from something brands attempt on their own to something a global holding company sells as a packaged service. Flywheel said the tool targets content gaps that shape how AI systems evaluate an item, including gifting context, age appropriateness, materials, and safety information, and that early pilots lifted product-page traffic and sales. The company has not published the methodology behind those results, and the figures are its own, so the verifiable news is the launch, not the performance claims attached to it.

Flywheel frames the stakes the way a company selling the service would. Mike O’Donnell, the agency’s head of AI, argued that products without the right content risk dropping out of AI-generated recommendations, which pushes brands toward paid media to hold visibility they used to earn through search. That is a sales argument, and it also points at a real tension suppliers should weigh on their own terms.

The Reason An Agency Is Selling This Now Sits In Walmart’s Earnings

Sparky has grown into a surface large enough to build a business around. Walmart told investors on its fourth-quarter fiscal 2026 call that roughly half of its app users had tried the assistant, and that customers who use it carry an average order value about 35 percent higher than those who do not. On the following quarter’s call in May, David Guggina, chief executive of Walmart U.S., said the number of items customers bought through Sparky more than quadrupled from the prior quarter. He described the assistant as helping Walmart “evolve from traditional search to intent-driven commerce.”

The surface also stopped being confined to Walmart’s own app. Beginning the week of March 25, after winding down its test of OpenAI’s Instant Checkout, Walmart embedded Sparky directly inside ChatGPT, and has said it will extend the assistant to Google’s Gemini. Walmart made roughly 200,000 products available in ChatGPT, and Daniel Danker, its executive vice president of AI acceleration, product and design, said the Sparky experience would travel into ChatGPT, Gemini, and other platforms Walmart integrates with. Discovery can now begin in an environment Walmart does not own and route back into a Walmart checkout, which is more places for a product to be named by an assistant, or skipped.

What These Services Work On Is Your Product Record

The reason a service like this has something to sell is that the input to an AI recommendation is the product record, and Walmart already treats that record as a discoverability lever. Walmart’s own Marketplace product detail policy states that an item’s title, description, key features, images, and attributes help determine how it appears in search and browse results, and that an item whose discoverability fields are left blank can be omitted from those results. Walmart describes Sparky as synthesizing reviews and making occasion-based recommendations, drawing on the same product record that policy treats as a discoverability input. How Sparky weighs that record internally, Walmart has not said, but the record is the shared input either way. Walmart even gives sellers its own AI-powered content suggestions for titles, descriptions, and key features inside Seller Center. The practical read for content teams is that the gap an agency offers to audit is one Walmart’s own tools already expose, and closing it does not require hiring anyone.

First-Party And Third-Party Sellers Face The Same Test Through Different Systems

The exposure runs across both segments, but the controls differ. First-party suppliers manage their item content through Walmart’s supplier item systems. Third-party sellers manage listings through Seller Center, where Walmart scores each item with a Listing Quality measure built from components including content and discoverability, offer, and ratings and reviews. A divergence worth flagging sits inside the third-party base: Walmart prioritizes content from brand owners and authorized resellers, so a reseller without those privileges cannot edit some attribute fields, including blank ones. The brand that owns its record can close a content gap directly, while a reseller may have to route the fix through the brand owner, a slower path to the same requirement.

The paid layer does not change that arithmetic, which is the part of O’Donnell’s pitch suppliers should test against their own data. Walmart has been testing ad formats inside Sparky, including sponsored prompts, since early this year, and Walmart Connect reported that 81 percent of surveyed shoppers had used Sparky to check availability or specifications before buying. A sponsored prompt still resolves into a recommendation pulled from the catalog, so buying the placement raises a question’s visibility without making a thin product record recommendable.

Walmart has said the assistant will keep expanding, into reordering and service booking, and eventually into reading image, audio, and video inputs. Each addition widens the range of product data the assistant can weigh, even as the criteria for being recommended stay undisclosed. An agency has now packaged a product to sell into that gap, which Walmart has left open by saying what Sparky will do next without saying how it decides.

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.