Case Study: Rakuten boosts product discoverability and click-through rates with ViSenze's Search by Image

Rakuten Uses Search by Image to Enhance Discovery and Engagement

Rakuten Ichiba, Japan’s largest e-commerce marketplace with tens of millions of users and millions of listings from about 40,000 merchants, faced a product discoverability problem: merchants’ descriptions often didn’t match how shoppers searched—especially in fashion. Rakuten set two goals: let shoppers find items that are visually relevant to their queries and enable discovery of complementary products by color.

To address this, Rakuten implemented ViSenze’s Search by Image so mobile users can search with photos or screenshots for visually similar items and filter recommendations by color. The AI/deep-learning solution improved product discoverability and click-through rates, with AWS infrastructure (EC2, EKS, Aurora, SQS, S3) providing the speed and accuracy needed.

Rakuten