Promotional sales make up a significant portion of revenue for fast moving consumer goods (FMCG) retailers. Consequently, accurate prediction of the quantities promotional sales in Konzum stores is important. TakeLab members were part of a team that developed a custom prediction model for Konzum.
Demand forecasting is paramount in supply chain management. Regular demand can be very accurately modeled using time series methods which consider historical data. However, when cosidering promotional demand, historical data can be very sparse, causing inaccurate forecasts. For many FCMG retailers, promotional sales make up a considerable portion of total revenue. Consequently, while difficult to achieve, improvements to such models can be very beneficial. TakeLab members were part of a team from Faculty of Electrical Engineering and Computing, University of Zagreb, that worked on improving the promotional demand forecasting system used by Konzum d.o.o.
Participants: Konzum d.o.o.; Faculty of Electrical Engineering and Computing, University of Zagreb
Duration: 1 years