Econometric Modelling and Forecasting of the Major FMCG Company Sales Volume

This study is devoted to statistical research and forecasting of the sales volume of the large FMCG company. With the help of cluster analysis, all regions of Russia were divided into two groups: regions with a highly developed and less developed retail market. Regression analysis in the selected clusters showed that the groups have different factors affecting the sales volume: in regions with less developed retail trade, the key determinant of sales growth is the increase in the distribution of goods, and for highly developed regions, the expansion of the product assortment was significant. In addition, the spatial interdependence of sales volumes between neighboring regions was found. Time series analysis allowed to build high-quality models of sales forecasting: Holt - Winters model and autoregressive model of distributed lags. Based on the results of the interpretation of the error correction model, it was concluded that the presence of a discount on the product increases the average sales in the short term, but in the long term leads to a decrease in sales. Thus, the studied interdependencies can be used to improve the quality of forecasting, as well as to create an effective marketing strategy of the FMCG company.