The growing popularity of the 15-min city (15MC) model has increased reliance on open geographic data to support proximity-based urban planning. However, the accuracy, completeness, and the categorization of such data remain underexplored and potentially problematic. This study examines the reliability of OpenStreetMap (OSM) and Google Maps in representing commercial amenities within the cities of Massa and Viareggio (Italy), using real-world data collected through extensive on-site surveys as a benchmark. The analysis addresses three core dimensions: the amount and spatial accuracy of digital data; the coherence and impact of categorization, assessed through datasets harmonization; and the implications for 15MC modelling. The findings reveal substantial discrepancies between digital datasets and ground-truth observations, underscoring critical limitations in the reliability of open data as informative strata. OSM presents substantial data gaps, while Google Maps shows better overall coverage but also a higher incidence of redundant or non-existent entries. The paper argues the need for accurate data validation and critical, context-aware use of open geographic data before engaging in spatial modelling – particularly when informing policy decisions within proximity-based planning frameworks – and concludes by discussing the methodological and practical implications for urban planners.