How Fanatics and OpenAI streamline the fan experience

Fanatics has launched an OpenAI-powered version of its Shop the Look feature across its network, including Fanatics.com and the official NFL, NBA, MLB and NHL online stores. When fans browse a team jersey, the product page now suggests coordinated outfits from in-stock items that match the team and style.

According to OpenAI's case study, the models draw on enriched product, team and inventory data, while Fanatics' merchandising rules keep picks relevant, and looks refresh daily. Testing across roughly 30,000 MLB products showed 10 percent higher engagement than the previous version.

This is a vendor case study, so we read it as OpenAI's account of a customer win, not independent results. With that said, it's a sensible use of AI. Matching hats to jerseys across a catalog that changes daily is tedious, rules-heavy work that scales badly by hand. Letting models assemble looks inside the retailer's own merchandising rules is the right division of labor.

The speed claim is the part we'd flag for other companies. OpenAI says the teams used Codex to build custom recommendation tools, tested the first 50 products in under 24 hours, and had a minimum viable product approved within the same week after a couple of days of iteration. If that holds up broadly, the barrier to building this kind of feature is dropping fast, and that's good for retailers who aren't Fanatics-sized.

The downside is modest but real. A 10 percent engagement lift is engagement, not sales, and the case study doesn't say how it translates into revenue or returns. And recommendation systems tuned to keep people clicking and adding to the cart are the same engines that push overspending. Sports merch is low stakes, but Fanatics also runs betting and gaming businesses, and the case study says it uses ChatGPT Work and Codex across those divisions too. Personalization techniques that feel harmless for outfits deserve much closer scrutiny when they meet wagering.

Our read: a tidy example of AI doing useful commerce work. I'm watching for whether the gains hold beyond engagement, and where else the same toolkit gets pointed.