Interchangeable or Inconsistent? A Comparative Analysis of Web Routing Services for Pedestrian Accessibility

ABSTRACT Web‐based routing services have become indispensable tools for accessibility research, yet their interchangeability is often assumed rather than verified. This study addresses this critical gap by conducting a systematic evaluation of pedestrian travel impedance estimates from three widely used platforms: Google Maps, ArcGIS, and OpenRouteService (ORS). Using 5000 origin–destination pairs connecting residential buildings to bus stops in Athens‐Clarke County, Georgia, this research quantifies platform agreement through statistical analysis and examines factors driving observed variations. The results reveal a fundamental paradox that while platforms demonstrate strong rank‐order consistency (Spearman's ρ > 0.97), statistically significant differences exist in absolute estimates. Travel time comparisons show pronounced divergence, with Google Maps generating estimates approximately 15%–16% higher at the median than ArcGIS and ORS. Distance estimates exhibit stronger agreement, with median differences of 1%–3% but mean differences reach 7%–9%. The findings demonstrate that while routing platforms are consistent for comparative analysis, their inherent algorithmic and data‐driven biases require researchers and practitioners to be critical consumers, as platform choice can significantly alter conclusions in fields ranging from urban planning and public health to social equity analysis.