<![CDATA[In the contemporary e-commerce ecosystem, delivery lead-time efficiency serves as a critical determinant of customer satisfaction. This study investigates the distributional characteristics of delivery durations for the "SPX Standard" service across four operational configurations: Early-Week Double-Day (peak), Weekend Double-Day (not available), Early-Week Non-Double-Day, and Weekend Non-Double-Day. Utilizing a quantitative comparative Exploratory Data Analysis (EDA) framework, historical logistics data were synthesized from 509 Shopee users in the Gunung Pati District, Semarang, spanning the 2024–2025 period. Statistical computation conducted via RStudio revealed significant departures from normality, as confirmed by Kolmogorov-Smirnov and Shapiro-Wilk tests (p<0.05). Consequently, hypothesis testing was performed using the non-parametric Kruskal-Wallis H-test. The analysis identified significant variance in median delivery velocities across groups (χ2=6.4256; p=0.04024). Findings indicate that while the Early-Week Non-Double-Day configuration recorded the highest median velocity (6 km/h), it exhibited extreme data dispersion, signaling process instability. Conversely, the Early-Week Double-Day period demonstrated superior reliability characterized by low variability, reflecting the efficacy of capacity scaling strategies during peak demand. The lowest performance metrics were observed during weekends, suggesting localized operational constraints. This research concludes that while peak-load operational stability effectively mitigates service degradation risks, standardized processes during regular workdays require further optimization. Limitations include the exclusion of holiday-related freight restrictions (e.g., Eid al-Fitr and Year-End periods).]]>
