With the emergence of new technologies that ensures information collection and effective information sharing between customers and O2O retail companies, new challenges arise to develop methods that can solve instant delivery problem of online orders in real time. In this paper, an online instant delivery scheduling model that considers the dynamic online order, decides the start time of deliveries and assigns riders to deliver online orders is proposed. The objective of the model is to minimize the return time and total distance of multiple deliveries. An online delivery scheduling algorithm (ODSA) is designed by introducing delivery rules with feasible delivery time windows. A case study of instant delivery for a retail store of an online-to-offline (O2O) retail company in Shanghai is explored to assess the effectiveness of the proposed model. The results show that with our proposed approach, 60% of orders can be delivered earlier than they would have been using the traditional delivery mode. Online scheduling improves the efficiency of instant delivery and enhances the competitiveness of the O2O company.