We study consumption profiling through the lens of attributed network clustering using proprietary bank transactions from eleven cities in Russia’s Far East (May–July 2025; 5, 000 observations per city). For each city, we construct an attributed network where edges reflect similarity in category-level expenditures (42 categories, online/offline) and node attributes summarize spending composition, temporal activity (morning/day/evening/night), mobility and loyalty indices, and age. This formulation yields coherent, city-specific consumer segments whose profiles align with socio-demographic and behavioral patterns, revealing consistent urban differences in essential and discretionary spending, online adoption, and mobility. Our contributions are a domain-tailored problem setup with a city-scale dataset and a comparative empirical study across representative attributed-network clustering paradigms, together with qualitative insights that support actionable segmentation for analytics, product design, and informed decision making. We conclude with directions for temporal tracking, learned similarities, and broader validation.