The emergence of online commerce has reshaped consumer behaviour, transitioning from traditional brick-and-mortar stores to digital platforms. This shift underscores logistical challenges, particularly in last-mile and first-mile reverse logistics, exacerbated by rising product returns. Scholars highlight the inefficiency of these logistics stages, driven by factors like underutilized vehicle capacity, parking troubles and low consolidation. Amongst academia, vehicle routing problems (VRPs) are a focal point of the optimization efforts. Notably, service time estimation is a critical yet often overlooked input of said problems, impacting routing accuracy and, thus, operational effectiveness. This study proposes a methodology for estimating service times in B2C last-mile and first-mile reverse logistics, validated through a real-world case study. Findings explore determinants of service time, and identify service point type and population density as significant factors.