This paper presents a methodological foundation for the Predictive Buying Intelligence Platform (PBIP), designed to address a critical inefficiency in merchandise planning within the luxury fashion industry. The proposed framework advances a comprehensive demand-forecasting approach that replaces traditional professional intuition with algorithmic analysis of 48 variables organized across five key dimensions: social signals (including TikTok virality), competitive intelligence, brand metrics, macroeconomic factors, and operational indicators. The document details a weighted scoring algorithm, a decision-making matrix, and a phased implementation protocol for embedding the system into existing retailer business processes. Particular attention is given to empirical validation, demonstrating forecast accuracy improvements up to 87% and a material increase in margin performance. The study is intended for retail top management, merchandising directors, buyers, merchandisers, planners and analysts focused on supply-chain digital transformation and the reduction of commercial risk.