Fanatics launches Shop the Look with OpenAI

Fanatics announced a new version of Shop the Look on October 26, powered by OpenAI models, that recommends coordinated apparel for game day. When browsing a team jersey, shoppers see suggestions on the product page for other available items that match the team and the style they’re looking for.

How outfits are put together

The models combine enriched product data with team and inventory information. Fanatics’ merchandising rules, the business criteria that guide item selection, keep recommendations compatible and relevant. Outfits are updated daily as the catalog changes.

The experience is available across many sites in the Fanatics network, including Fanatics.com and the official online stores of the NFL, NBA, MLB, and NHL. OpenAI describes how Shop the Look works and where it’s available.

From initial test to production

OpenAI and Fanatics’ engineering team used Codex to develop recommendation tools. These enrich catalog data, find compatible items, evaluate recommendations, and generate outfits for each product. The initial test with 50 products was completed in less than 24 hours.

Fanatics reviewed the outfits and gave feedback directly to the people developing the tools. After a few days of adjustments, it approved a minimum viable product, an initial functional version, that same week. OpenAI helped design the solution, and Fanatics implemented the experience in its production environment and expanded the approach to thousands of products.

The reported result for MLB

According to OpenAI, after testing recommendations across approximately 30,000 MLB products, the new experience saw 10% more interaction, called engagement in the announcement, than the previous version of Shop the Look.