For a selected range of jeans (22 items), which is one of the most challenging categories in terms of finding the right size due to the lack of sizing standards, customers can see how different sizes from various brands would fit them, with a heatmap indicating where the item sits tight or loose on the avatar they created, Zalando said in a press release.

Zalando has already run two pilots with selected items of clothing from Puma and Zalando’s private label Anna Field, where more than 30,000 customers tried out the technology feature. These pilots run for several weeks at a time.

Zalando’s team uses a combination of fitting models, machine learning, computer vision, and other technologies to predict if items will run big or small. It has also created personalised size recommendations based on customers’ purchase and return history along with reference items that customers can add to their size profile. For those items where Zalando provides size advice, size-related returns have decreased by 10 per cent versus similar items where size advice is not provided.

Zalando’s Size and Fit team is also working in parallel on a body measurement feature, which will allow customers to receive personalised advice based on their actual measurements. This will be rolled out in the upcoming months.

“Our goal with these pilot campaigns is to learn and understand how customers engage with this new technology so we can develop a seamless scalable solution for the future. We can already see that customer engagement with these campaigns increases and, in fact, around half of the customers try more than one size on the avatar,” explained Stacia Carr, vice president of Size and Fit at Zalando. “We want to support the industry as it continues to adopt and leverage 3D digital design software and workflows to produce fashion. These processes generate digital assets necessary to scale a virtual try-on experience.”

Fibre2Fashion News Desk (NB)