As vital components of urban commercial infrastructure, retail facilities serve as primary carriers of urban commercial functions. Achieving differentiated and equitable retail layouts is critical to address spatiotemporal shifts in consumer demand and promote sustainable development. Existing retail location models often insufficiently integrate complex factors like consumer demand heterogeneity, urban spatial structures, and facility hierarchies. This study chooses four Chinese national-level new areas including Xiong'an, Binhai, Pudong, and Nansha as study areas to represent distinct urbanization stages. Leveraging multi-source geospatial data (POI, road networks, population, GDP, nighttime lights), we propose a recursive hierarchical optimization method. The approach quantifies power-law relationships between retail tiers and consumer density through rank-size curves, coupled with an alternating process of excluding higher-level facility coverage and optimizing maximum coverage for lower-tier facilities. The results show: (1) Post-optimization hierarchical networks (8, 11, 13, and 16 tiers) minimized "retail deserts," reducing uncovered areas to 3.15 %, 0.59 %, 3.93 %, and 5.62 % respectively. (2) Enhanced 15-min pedestrian and driving accessibility revealed polycentric spatial synergies and expanded coverage. (3) Urbanization levels significantly influence demand and layouts: rapidly urbanizing areas balance service upgrades with coverage expansion, while developing regions prioritize proximity-based convenience facilities. The recursive hierarchical method effectively aligns facility placement with demand heterogeneity, offering a scalable framework for consumer-centric retail planning and sustainable urban development.