Successfully Improving Cart Conversion Rate by 10%
with AI Semantic Search Model at Kurly
AI startup Dalpa announced on July 15 that it has provided its AI search solution to the e-commerce platform Kurly, successfully addressing the issue of "no-result search."
During product searches, issues such as typos, omission of synonyms, and a variety of natural language expressions can prevent search queries from matching the product database, resulting in failed searches. This has been identified as a major cause of customer drop-off during the purchase process.
To address these challenges, Dalpa designed a semantic search model for Kurly that precisely analyzes the meaning of search queries and recommends appropriate products. This model is an AI-based search engine specialized in processing natural language, typos, and unstructured queries.
To verify the effectiveness of the new search AI, the two companies conducted a test comparing Kurly's existing search system with Dalpa's model. As a result, the group using Dalpa's search AI saw a roughly 10% increase in cart conversion rate. This improvement in search quality alone, without any additional promotions, led to a tangible increase in sales.
In terms of cost, Dalpa's solution demonstrated advantages such as lower implementation and operational costs compared to existing search systems, as well as rapid customization. This enabled both enhanced search quality and reduced operating expenses.
Park Hyoyeon, leader of Dalpa's commerce team, stated, "This collaboration is a representative case in which AI technology has delivered concrete results based on actual customer behavior data," adding, "In e-commerce environments like Kurly, where the search experience is critical, this project has once again proven that precise search quality directly translates to higher conversion rates."
© The Asia Business Daily(www.asiae.co.kr). All rights reserved.
