mySHOEFITTER reduces size-related returns by more than 90% with AI sizing

Buying shoes online leaves one question unanswered until the box arrives: will this size fit? mySHOEFITTER answers that question before checkout. A shopper takes one smartphone photo, completes five clicks, and receives a size recommendation in about 30 seconds.

The product addresses the part of the returns problem that sizing can solve. mySHOEFITTER reports that its customers have reduced size-related shoe returns by more than 90 percent after adding the sizing flow to their stores.

Research translated into a simple shopping flow

Fraunhofer FIT initiated the joint venture behind mySHOEFITTER. Its researchers worked with the product team on biometric foot measurement for mobile devices, combining computer vision and AI to match measured feet with the dimensions of a specific shoe model.

The early prototypes still deviated by 11.2 millimeters. After thousands of development hours and another 100,000 measured and scanned feet, mySHOEFITTER reports millimeter-level measurement accuracy. Shoppers only see the short photo flow; the measurement work happens behind that simple interface.

LLMBase’s role

We contributed AI technology and product development to the joint venture. Our work helped turn the research into a flow that retailers can place inside an online store and customers can complete without special equipment.

Fraunhofer FIT contributed biometric measurement research, mySHOEFITTER built the retail product, and we helped deliver the AI experience. mySHOEFITTER reports the customer outcome directly: fewer returns caused by choosing the wrong size.