Korea's Delivery Apps Turn to AI to Predict Arrival Times
South Korean delivery apps Baedal Minjok and Coupang Eats are deploying AI to enhance service quality. Baemin will implement an AI system in October to predict delivery times, using 58 data types. This system has improved forecast accuracy to over 50%, a 1.7-fold increase over manual estimates. Coupang Eats has developed store-coordinate optimization technology, reducing inquiries about restaurant locations and cutting delivery times exceeding one hour.
AI for efficiency is one thing, but this is about customer retention in a cutthroat market. Baemin's 1.7-fold accuracy improvement over manual estimates directly addresses a key customer pain point. Coupang Eats' coordinate optimization similarly tackles rider frustration and late deliveries. This is a defensive play against churn, not just an innovation push.
South Korea's delivery market shows intense competition, with Baemin's monthly active users at 24.4 million and Coupang Eats growing 23.0% to 14.44 million. This AI investment reflects a battle for repeat business. Companies like Grab in Southeast Asia, facing similar competitive pressures, will watch these developments closely. The focus on reducing delivery uncertainty could become a regional benchmark for customer satisfaction.
The thing to watch is whether these AI improvements translate into sustained user growth and market share shifts. Baemin's 5.8% user growth lags Coupang Eats' 23.0% increase over the past year. If Baemin's AI system significantly closes this gap by early 2027, it will validate the investment. Otherwise, it is just table stakes for competition.
Related reading
6 stories
Chinese regulator drafts rules to protect teenagers from ‘intimate’ AI companions

US, China agree AI dialogue ahead of Trump-Xi summit

Neither the US nor China can win the AI race alone, BofA analyst says

StashAway Becomes First Investment Platform to Join yuu Rewards Club

Trust Freedom Card Adds Miles Rewards With KrisFlyer and HeyMax

