Introduced “aigent recommender” as a recommendation engine, and monthly EC sales via recommendations reached 3~5 million yen.

【 Summary 】

Kyoto Kimono Ichiba corp. is one of Japan’s largest “kimono” Internet shopping sites. It is divided into three businesses: exhibition sales, stores, and EC, with 50% of sales coming from exhibitions, and the remaining 25% each from stores and EC.
In the EC business, the company is expanding not only its EC functions but also its telephone support system to ensure that customers are satisfied with their purchases, even if the purchase process is lengthy because they are expensive purchases. Since many of the customers are relatively elderly, abundant and easy-to-understand product descriptions and a stress-free purchase path are very important, and the enhancement of product search and the introduction of a recommendation engine was also necessary to realize a stress-free shopping experience.
The reason for choosing “aigent recommender” as the recommendation engine this time was its AI-based learning effect. The company thought it would be cost-effective because it would automatically optimize the product without the hassle of operating it. However, since many of the products are expensive one-of-a-kind items, we knew that buying them in combination would not be effective, so we tuned the algorithm based on the browsing history of the products and used it as a guideline to encourage customers who are considering purchasing to compare products and consider combinations.
After the service was introduced, the company reported monthly sales of 3-5 million yen as sales via recommendations. Customers feel that they do not want to be pushed to buy expensive kimonos, so instead of forcibly recommending hot-selling products, they are happy to encounter products that they can like with less unmatching. The aigent recommender is a tool that can be used in line with this philosophy.