Fanatics and OpenAI integration for Shop the Look

Overview of the Shop the Look integration

Fanatics has deployed an OpenAI-powered version of its 'Shop the Look' feature, which utilizes AI models to suggest coordinated merchandise for sports fans. The tool is designed to help users complete their game-day outfits by recommending compatible items based on the specific product they are currently viewing.

Technical implementation and data processing

The 'Shop the Look' feature functions by synthesizing enriched product data, team-specific information, and real-time inventory levels. OpenAI models process this data to assemble relevant outfit combinations, while Fanatics applies internal merchandising rules to ensure that all recommendations remain stylistically compatible and relevant. These recommendations are refreshed on a daily basis to account for changes in the product catalog.

Development velocity and engineering impact

Fanatics utilized OpenAI's Codex to accelerate the development of custom recommendation tools. The engineering team was able to build, test, and iterate on the initial 50 product looks within 24 hours. Following a rapid feedback loop, the company moved from the initial prototype to an approved minimum viable product within a single week.

Performance metrics and scale

The implementation of this AI-driven approach has allowed Fanatics to scale recommendation generation across thousands of products. In a test involving approximately 30,000 MLB products, the new 'Shop the Look' experience resulted in a 10% increase in user engagement compared to the previous recommendation system. Beyond this specific feature, Fanatics utilizes ChatGPT Work and Codex across its broader corporate, commerce, collectibles, and betting divisions.