Kolon Mall wanted to help shoppers find clothes that truly matched their personal taste. But their old recommendation system couldn’t recognize how fashion items were similar in design. It only looked at simple similarities, such as the same brand or price points. Completely missed small but important details — silhouette, pattern, vibe, occasion — that made items feel alike. Because of this, even with a huge, growing catalog, Kolon Mall struggled to suggest products that truly felt right for each shopper.

That changed when the team started adopting YesPlz AI’s style-focused recommendation engine. The AI continuously updates which design elements each shopper is drawn to, then suggests other pieces with a similar look and feel. The results came fast. In just two weeks, sales went up 15%. Checkout conversion rates increased by 13.5%. And average cart value grew by 10%, as shoppers found more fashion items that matched their unique sense of style.

Challenges

• Rule-based recommendations without style matching: Kolon Mall featured several types of product recommendations on its detail pages, but none were driven by style. The system relied on simple rules, suggesting items from the same brand or category rather than those with a similar look or feel. Recommendations, therefore, felt random and impersonal, and shoppers quickly lost interest.

• Shoppers couldn’t find what they wanted, even with hundreds of products added daily: While the catalog kept increasing, the product data behind it didn’t. Manually tagging products was slow and inconsistent. What made each item special — the cut, fabric, or any design details — was often missed. Without rich, accurate tags, the recommendation engine couldn’t connect shoppers with the products they’d actually love. Finding the right item thus felt harder as the catalog grew. 

• Shoppers preferred brands they already knew, so hidden gems were overlooked: Many stayed loyal to big, well-known brands. Private labels and smaller fashion brands didn’t get much attention, even though they had great designs and value. To address this brand bias, the Kolon Mall team wanted to leverage each shopper’s history for personalized recommendations. However, technically handling many aspects of a shopping history to reflect individual preferences accurately was not that simple.

Key Results:

• 15% increase in sales within two weeks of adopting AI-driven recommendations.

• 13.5% rise in checkout conversion rates within just two weeks.

• 10% growth in average cart value

• 4 YesPlz AI Tools Leveraged

We’ve immediately experienced a higher conversion rate with a significant increase in page views. Most of all, we like how the YesPlz team is there to make sure the integration is easy.

Gyyoung Kim

CGO, Kolon Industries FnC

Solutions: 

• Style-focused recommendations that match shoppers’ tastes: YesPlz AI looks at the design elements of each item and how shoppers interact with them. Based on those insights, it figures out which features matter most to each shopper. It then recommends other items with similar styles. Shoppers get suggestions that actually match their personal taste.

• Customized AI recommendations to Kolon Mall’s unique collection: The AI quickly adapted to Kolon Mall’s mix of brands and unique styles. The fine-tuning enabled it to understand what Kolon Mall fashion-savvy shoppers wanted at a micro level, making recommendations more accurate, relevant, and personal. 

• Solving cold-start challenges with style-based curation: Traditional recommendation systems often struggle with cold-start problems—when new or lesser-known products lack enough data to be recommended. YesPlz AI’s style-based engine solves this by focusing on visual and design attributes rather than user behavior alone. At Kolon Mall, this approach made it possible to surface hidden gems, older inventory, and emerging brands that matched shoppers’ style preferences, increasing both product discovery and sales opportunities.
  
  AI-Powered Curation: Before and After

Before

After

AI-Powered Curation: Before and After

Before

After

Personalized Curation: Before and After AI

Before

After

Lesson Learned

Kolon Mall thought they had recommendations covered with four types already live on product detail pages.

But adding YesPlz's style-based recommendations changed everything—an immediate sales lift of 15% kicked in within two weeks. Even buried, older inventory started selling as shoppers discovered hidden gems that matched their vibe.​

We also saw collaborative filtering work in perfect synergy, helping solve the cold-start problem when presented alongside style-based recommendations

Real growth happens when you layer AI that understands your unique assortment and shopper preferences, instead of relying on generic recommendation engines..

Behind the Curtain

• Kolon Mall case study: How Style-Driven AI Sparked a 15% Sales Surge in Two Weeks