Can AI-powered demand forecasting fix fashion’s inventory crisis?

Fashion has an inventory management problem, and it’s only intensified in recent years. Solving these challenges has proven elusive, leaving the industry with billions in unsold stock annually and fuelling a system running on baked-in excess where margin-killing markdowns are the norm.

This is largely due to consumers demanding faster, trend-driven cycles, which reduce lead times and increase the need for brands to accurately forecast demand. Global disruptions, ranging from supply chain delays to unpredictable shifts in consumer behaviour, which stem from events like the pandemic, have also exacerbated forecasting challenges. Traditional demand patterns are no longer reliable, as external factors like unexpected lockdowns, fluctuating consumer spending and the climate crisis have injected further unpredictability. Seasonal volatility has also grown, with more extreme weather events impacting everything from materials sourcing to shipping timelines, making it harder to maintain accurate forecasts.