Accurately predicted demand surges for leading online retailer
Case Studies / Retail / Price Optimization
A large online cosmetic retailer was unable to accurately predict demand. They routinely experienced “out of stock”’ conditions, resulting in lost sales. To compensate, they were forced to maintain high inventory levels at all times, which was unsustainable (and costly).
Our goal was to create a forecasting model that would accurately predict demand by taking into account the effects of campaigns, the impact of price changes, and consumer preferences for similar products.
We combined customer transactions, online behavior, campaigns, cross-product elasticity, and substitution effects to create an advanced forecasting model.
Our forecasting model helped the client anticipate demand changes. This allowed them to purchase product at opportune times and reduce their inventory maintenance costs. Most importantly, our model was 15-20% more accurate than existing models the company had used.