International outdoor brand reduces demand forecast error by 55% using AI
Challenge: Balancing supply and demand at scale
Each season, our client’s forecasting team faced the complex task of determining how much of each product to manufacture up to three seasons in advance. Relying on a mix of fundamental analytics and intuition often proved inefficient and prone to error. This resulted in frequent mismatches between supply and demand. Some products were overproduced, tying up valuable resources and creating excess inventory, while others were underproduced, leading to stockouts and missed revenue opportunities.
The consequences extended beyond operational inefficiencies. Inaccurate forecasts drove up costs, created unnecessary waste, and complicated planning across production, logistics, and sales. Recognizing the growing importance of more reliable, data-driven forecasting, the client engaged us to design a solution that would improve demand predictability and enable smarter, more efficient resource allocation.
How we helped
How can we forecast inventory needs for new items when we have limited data?
To address the “cold start” problem, we leveraged the characteristics of upcoming products, including colors, categories, and price points. It allowed us to replace gut-feeling predictions with accurate ones.
We developed a solution to forecast demand at the product and store levels, using historical stock and external variables such as calendar effects and seasonality. This process involved:
- Analyzing color clusters to predict future color success
- Forecasting product reorders and preorders
- Integrating experts' input to improve forecasting accuracy
The impact: Transforming forecasting into a strategic advantage
- Reduced the forecasting error by 55% resulting in 16M/year in business benefits.
- With better preorder predictions, the client reduced reorders and logistical costs associated with airplane shipping, driving sustainability.
- The access to extra insights, integrated into product development & design, won the client a Group Level Innovation Award.
