Walmart Pivots From Instant Checkout to Sparky on ChatGPT

One of the world's largest retailers tried selling products directly inside ChatGPT, and the results were disappointing. A Walmart executive told WIRED that sales through Instant Checkout, the feature that let ChatGPT users complete purchases without leaving the chat, were "disappointing," citing wrong items appearing in carts and conversion rates below Walmart's own digital channels.

Walmart's response is a strategic pivot. Instead of relying on OpenAI's checkout flow, the retailer is embedding its own Sparky chatbot directly into ChatGPT and Google Gemini. Sparky, live on Walmart.com and the mobile app since early 2025, is built on Walmart's proprietary retail data and third-party language models. The Sparky-in-ChatGPT rollout begins this week for ChatGPT Plus and Gemini Advanced subscribers, with free tier access coming later this spring.

Why Instant Checkout underperformed

The Walmart-OpenAI partnership launched in October 2025, allowing select ChatGPT users to purchase Walmart products through Instant Checkout. Two problems quickly emerged: inaccurate product results pushed wrong items into user carts, and conversion rates fell below what Walmart sees on its own app and website.

This pattern mirrors OpenAI's broader retreat from native in-chat checkout. OpenAI confirmed to Modern Retail that Instant Checkout is "moving to Apps, where purchases can happen more seamlessly," a quiet acknowledgment that users are not yet completing purchases inside AI conversations at scale. Following this shift, Shopify says it will allow all stores to appear inside ChatGPT through is agentic storefronts, with purchase redirected to merchant sites.

The Sparky strategy: brand-controlled AI commerce

Walmart's pivot follows a logic that Amazon has applied with Rufus: run discovery and research through the AI platform, but keep the purchase experience on a branded, retailer-controlled surface. Sparky-in-ChatGPT routes users to Walmart.com or the mobile app to complete transactions rather than finishing inside the chat.

This matters beyond Walmart. The lesson from the Instant Checkout experiment is that trust and familiarity drive checkout conversion. Consumers who already have Walmart accounts, saved addresses, and payment methods complete purchases on Walmart's own surfaces, not inside an AI interface they are still learning to use.

What it means for AI commerce strategy

Walmart's pivot and Amazon's Rufus both point toward the same model: AI platforms are discovery and research environments; owned channels are conversion environments. For brands in retail media, the implication is that AI commerce investment should focus on product visibility in AI search results, not on in-chat transaction completion.

The brands that win in AI commerce are likely those that optimize their product data for AI discovery, making sure their items surface accurately in Sparky, Rufus, and similar tools, rather than those betting on frictionless in-chat checkout as a near-term behavior.