GermanyRetail and e-commerce2018 – ongoingZalando
Machine-learning size recommendations
Zalando reported 10% fewer size-related returns on items with size advice
Problem
Clothing sizes varied across brands, making online fit difficult to judge. Poor fit led to returns.
Approach
Zalando built machine-learning systems to predict how products fit. Personalized recommendations used purchase and return history alongside customers' reference items.
Results
Takeaway
The reported comparison concerned items with size advice versus similar items without it. It did not establish the effect of the separate avatar fitting-room pilot.
Categories
- Size
- Large enterprise
- Industry
- Retail and e-commerce
- Business function
- Marketing and sales
- Innovation type
- Product and service innovation
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Zalando described its size-recommendation systems alongside a separate virtual fitting-room pilot. It reported 10% fewer size-related returns for items receiving size advice.
Zalando dated its machine-learning Size & Fit work to 2018. It explained how product information and customer feedback supported recommendations.
Full details are in the original sources.
