One international cosmetics retailer frequently misaligned stock levels with consumer demand, which was causing serious problems with revenue and margin.
Unable to accurately predict purchase volumes in various territories, and in physical stores vs online, the retailer turned to Agile to help them forecast demand and carry appropriate stock levels in the relevant warehouses and stores.
Here is how we:
- Designed a Machine Learning model to predict stock requirements
- Trained the model free of unwanted bias
- Tested the model
- Gave buying teams confidence that they would not understock or overstock specific stores or countries