AI기반 묶음배달의 효율성이 라이더 수입에 미치는 영향

Abstract

The outsourcing of food delivery began in the late 1990s, and the growth of food delivery platforms provided consumers with the opportunity to access a variety of foods through delivery. However, due to the nature of the restaurant industry, food must be cooked upon ordering and delivered to the consumer in a warm state. This takes less than an hour, which makes it difficult for restaurants to expand their delivery service. Therefore, only 20.9% of total restaurant sales are generated through food delivery via phone orders and food delivery platforms, and sales through food delivery platforms are expected to be even lower. This outsourcing of food delivery is similar to changes in the logistics industry due to the development of e-commerce. The logistics industry has achieved economies of scale through a hub-and-spoke logistics system and made faster delivery to consumers a competitive advantage. Faster delivery has emerged as a major competitive factor in food delivery as well. However, unlike the general logistics industry, there is a problem in achieving economies of scale. To address these challenges, each food delivery platform invested in various technological developments. But with the high inflation rate continuing after COVID-19, consumer needs have changed. In response to these environmental changes, Baedal Minjok has launched an AI-based STOD(Stacking Owned Delivery) service that combines multiple orders into one delivery, thereby shifting away from the existing speed-centered competition. However, these changes raise questions about the effect that the launch of the AI-based STOD(Stacking Owned Delivery) service will have on earnings, as delivery riders perceive a high degree of uncertainty about their earnings. Therefore, in this study, we use the difference-in-difference method as a quasi-experimental design method, based on data from Woowa Youth, the operator of Baemin Connect, to examine the impact of AI-based STOD on efficiency and rider earnings. As a result, the monthly earnings of riders who used both AI-based STOD and single-house delivery increased by 5.1% compared to delivery riders who only used single-house delivery, and earnings per one delivery increased by 0.8%. Additionally, hourly earnings increased by 4.9%, demonstrating that the launch of STOD contributed to the increase in delivery rider earnings. Delivery rider earnings also increased similarly across regions. Based on these research results, we discuss the implications and limitations of this study.