Retail's AI Future Built in Arkansas Warehouse

A large warehouse in Bentonville, Arkansas, serves as Eko's "capture factory," where hundreds of employees meticulously create digital product catalogs. This manual process involves photographing items like vitamin bottles and strollers from every angle, adjusting lighting, and ensuring pristine surfaces to build AI-ready product files. The goal is to significantly improve the accuracy of online listings for retailers such as Walmart and Best Buy, making product information easily digestible for artificial intelligence platforms.

Eko's effort is central to the emerging AI-driven shopping landscape, where chatbots from OpenAI and Google are increasingly used for purchases. As AI platforms delve into product specifics, details like shoe width or precise descriptions become paramount. Retailers, long challenged by inaccurate product listings often outsourced to suppliers, are now re-evaluating their inventory data to be AI-ready.

Walmart has invested over $300 million in Eko since 2018. Eko retains ownership of the product information, leasing "Eko files" to brands and retailers, which prioritizes accuracy. For instance, Eko insists on realistic images, even if a product appears less fluffy than a seller desires, aiming to reduce returns and boost sales through accurate representation. The creation of an "Eko file" can range from ten minutes for simple items to half a day for complex ones like refrigerators.

This venture highlights the dynamic tension between retailers and AI companies as they shape the future of online shopping. While AI shopping is still developing, companies like Eko are providing the foundational data. OpenAI's recent pivot from its Instant Checkout feature to directing users to retailers preserves retailer-shopper relationships, underscoring the complex evolution of e-commerce.