A Review of Fast Fashion Demand Forecasting Methods

Abstract

Fast-fashion retail operates in an environment characterized by short product life cycles, high demand volatility, and rapidly changing consumer preferences. Accurate demand forecasting is therefore essential for reducing inventory risk and improving suppl y chain responsiveness. This paper provides a comprehensive review of demand forecasting methods applied in fast fashion. The objective of this review is to analyse the existing traditional forecasting approaches, machine learning techniques, and hybrid forecasting systems. Different specific models were chosen based on the three categories, such as ARIMA, Regression-based for traditional models, LSTM, social media-driven methods for artificial intelligence-based models and hybrid models which combined elements of both traditional and AI-based. The paper concludes that no single method dominates across all fast-fashion forecasting; however, different types of models are used based on specific characteristics. Hybrid methods are considered more promising than traditional forecasting methods due to the reviewed literature showing overall best performance.