Loading...
USING AI-POWERED DEMAND FORECASTING TO REDUCE OVERSTOCK AND WASTE IN FAST FASHIONE-COMMERCE
Ba Bakr, Norah
Ba Bakr, Norah
Citations
Altmetric:
Type
Supervisor
Date
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
This study examines the role of AI-powered demand forecasting in fast-fashion e-commerce,
assessing its capacity to reduce overstock and waste while improving customer satisfaction. Using
a mixed-methods approach — a PRISMA-guided systematic literature review of global case
studies (Zara, H&M, Shein, Amazon, Target) and a quantitative survey of 72 participants involved
in operations, logistics, courier services, fast-fashion firms, and OSCM students — the research
maps technological benefits and contextual gaps relevant to the Saudi market. Findings show that
AI-driven models (particularly neural networks and multi-source predictive systems) substantially
outperform traditional techniques by incorporating real-time signals such as social media trends,
promotions, store traffic, and seasonality. The paper concludes by advocating for regionally
calibrated datasets, capacity building, public–private collaboration, and hybrid human–AI
workflows to increase adoption and reliability. Implications extend to Saudi fast-fashion e-tailers
seeking to align inventory strategy with cultural and seasonal demand patterns (e.g., Ramadan,
Eid), while minimizing environmental and financial costs.