Configuration of Sustainable Distribution Networks as a Determinant of Logistics Coordination Mechanism Selection

This study’s purpose was to analyze how the configuration of a sustainable distribution network affects the effectiveness of logistics coordination mechanisms, specifically focusing on the role of 3PL (third-party logistics) providers. We examined 69 networks that used a 3PL provider. The study used a weighted regression approach, with coordination mechanisms scaled by importance using the Analytic Hierarchy Process (AHP). To enhance interpretability, the SHAP model from Explainable AI (XAI) was used to identify the most influential configuration factors, which included service recipient type, product characteristics, warehousing susceptibility, and assortment diversity. The findings indicate that while increasing network complexity enhances adaptability, it may simultaneously reduce the efficiency of certain coordination mechanisms. The study highlights warehousing susceptibility as a critical factor, with other variables having a weaker or statistically insignificant effect. The SHAP analysis provided additional practical insights beyond standard statistical thresholds. By integrating expert-based weighting (AHP) with XAI, we propose a hybrid analytical framework that helps 3PL operators select the most effective coordination tools, such as flow forecasting, for specific network and product types.