Fit and Virtual Try-On

I partnered closely with adidas’s Future team, the product innovation group, on joint research, pooling our budgets to go further. Future focused on product creation, which runs on long timelines. My team tested digital solutions we could validate in six to twelve months, and our findings fed back into their product conversations. We evaluated dozens of technologies and ran multiple pilots to find what gave shoppers the confidence to buy online.

Technology from the Virtual Showroom led to a pilot to digitize footwear: scanning shoes for Virtual Try-On in the adidas app. It has since scaled to most footwear models, using 3D photogrammetry scans captured at the product photography studio. Footwear Virtual Try-On lifts conversion 50% in the app.

The MySizeID pilot aimed to reduce returns by giving shoppers confidence that apparel would fit. MySizeID, a machine-learning startup, builds a “virtual twin” of a person’s body to predict a size. We ran two pilots in Portland with 150 participants in total: first at the corporate office with interns and employees, then at the employee store, with shoppers invited through the adidas app. Shoppers picked the size they thought would fit online, then used MySizeID, and both sizes were tried on in a dressing room. 90% of female shoppers picked the wrong size online. With MySizeID, that dropped to 40%.

Early on, the sizing tool required its own app. Today similar solutions plug directly into ecommerce sites, which makes them much easier to adopt. The sizing work lost global roadmap priority to tech debt and new-market expansion, while footwear Virtual Try-On scaled.