Case Study: Neowing boosts AOV and long-tail product discovery with Algonomy DeepRecs NLP

Game-changing DeepRecs NLP Improves Product Discovery of 2M+ Catalog with 96% Long-tail Items

Neowing, a Japanese online shop (and its international sister site CDJapan) sells over 2 million entertainment products, but more than 96% of the catalog is long-tail with little behavioral data, making relevant recommendations and product discovery a major challenge. To address this, Neowing engaged Algonomy, using Algonomy DeepRecs NLP (along with Recommend™, Discover™ and Engage™ capabilities) to generate recommendations from Japanese product descriptions rather than relying on past purchase or browsing history.

Algonomy implemented DeepRecs NLP across item pages and validated the approach with Algonomy Experience Optimizer (XO), surfacing relevant new and long-tail products. The results included a 6.25% lift in Average Order Value and a 4.99% CTR lift on neowing.co.jp, and on cdjapan.co.jp a 7.93% lift in Revenue Per Visit, 3.55% higher conversions, a 4.54% CTR lift and 8.39% more units per order — demonstrating Algonomy’s measurable impact on discovery, engagement and revenue.

Neowing

Katagiri Fumio

Chieff Executive Officer