Predicting customer demand is vital for large-scale retailers to optimize supply chains and to support business decision-making. This is especially true in FMCG, where products have a short shelf life and the costs of oversupply and over manufacture can be significant. As new products emerge, stores open in new locations and consuming patterns change, the demand forecasting problem becomes ever more complex.

Demand forecasting is a complex problem that requires thorough analysis and robust solutions. Various factors such as seasonality are of influence. At ML6, we developed and delivered time-series forecasting models in numerous projects. we have extensive experience with processing the required data and extracting meaningful features that can be used together with historical data to build demand forecasting tools.