Musinsa said on the 12th it will introduce an "AI review summary" service that uses artificial intelligence (AI) technology to grasp tens of thousands of product reviews at a glance.
The service was planned to provide a more efficient fashion shopping experience by having AI directly analyze and refine about 21 million real-user review data points accumulated in the Musinsa Store over the past two years.
The service provides AI-analyzed and refined real-user review data of about 21 million entries accumulated in the Musinsa Store over the past two years. Musinsa has a structure in which about 11.52 million reviews are generated annually, averaging 960,000 per month. To help customers quickly grasp key information such as fit, material, and wearability amid the vast data, a summary function was placed at the top of the review section on the product detail page (PDP).
The service reflects the characteristics of the fashion category. Through the "pros and cons summary" function, it presents not only positive evaluations but also noteworthy drawbacks, enhancing the objectivity of the information.
It also provides category-specific custom keywords. For example, pants are summarized around "size, fit, stretch," and outerwear around "design, warmth, material."
Musinsa also introduced a feedback function that allows users to directly rate the AI summary as "Helpful" or "Not helpful." The collected ratings will be used to further refine sentence generation and improve the accuracy of the AI model.
A Musinsa official said, "In the five days after the service launched, about 80% of customers clicked 'Helpful,' confirming its practical utility," adding, "It is currently in beta testing, and we plan to continue advancing the model based on user feedback and expand AI features that improve shopping convenience."