← All case studies

Fraunhofer FIT

Fraunhofer FIT initiates the first AI flow to precisely determine shoe sizes

Fraunhofer FIT was the main initiator of a joint venture between several companies and LLMBase to build the first AI flow for precisely determining shoe sizes.

20-30%

Previous return rate

Typical shoe return rate before the digital fitting flow.

2-3%

Return rate after fitting

Observed return rate after customers used the fitting flow.

>90%

Reduction in size-related returns

Drop in fit-driven returns compared with the previous baseline.

Fraunhofer FIT was not just a customer context for mySHOEFITTER. It was the main initiator of a joint venture between several companies and LLMBase, focused on one practical e-commerce problem: shoe returns caused by uncertain sizing.

The venture built the first AI flow that could precisely determine shoe sizes from customer input. The team behind LLMBase contributed to the digital fitting experience and kept the implementation grounded in measurable retail outcomes rather than generic personalization claims.

The fitting flow helped users capture the information needed for a better size recommendation before checkout. That shifted the purchase decision from estimation to a guided fit process and gave the project partners a clearer way to evaluate whether the experience changed return behavior.

Sweetwater Marketing added partner context around the project and its market positioning, including the Adidas partner background supplied for this case study. The public mySHOEFITTER materials remain the source anchors for product context, company background, and customer-facing FAQ claims.

The result was a sharp reduction in size-related returns: from a previous 20-30% return rate to 2-3% after customers used the fitting flow.