Optimizing E-Commerce Logistics Experience Through Information-Driven Intervention

This paper develops and empirically evaluates an information-driven intervention framework for optimizing end-to-end customer experience in e-commerce logistics. It is motivated by the diminishing returns of competition focused solely on delivery speed and the absence of a closed-loop mechanism linking logistics actions, customer emotions, and economic value. The study integrates multi-source operational, trajectory, and review data; constructs a five-dimensional experience indicator system; and applies a Light Gradient Boosting Machine (LightGBM)–Text Convolutional Neural Network (TextCNN) attention model within an online A/B testing scheme using large-scale platform orders. Results demonstrate significantly higher customer satisfaction, compression of long-delay tails, improvements in on-time performance and information visualization, and measurable gains in repurchase rates and revenue under latency constraints. These findings indicate that calibrated transparency and targeted data-driven interventions outperform indiscriminate time compression, offering a scalable blueprint for experience-driven logistics operations, data-centric service innovation, and intelligent logistics planning.