Enhancing demand forecasting: an analysis of factors impacting sales and implementation of improved methodologies for accurate prediction

Demand forecasting is very crucial for companies in the FMCG sector especially in today’s dynamic market environment. The thesis explores the complicated nature of forecasting in an FMCG company, encompassing both long-term and short-term techniques. Long term forecasting works through carefully purging remaining inputs from other departments and a thorough analysis on sales data from previous year. These inputs are carefully calculated and incorporated into future projections with exactitude. Contrary to this, short term forecasting is traditionally carried out through dividing out of the long-term forecast while accounting for weekly fluctuations. However, this thesis presents a more intricate approach to short-term forecasting. This implies that it seeks to bring together long-term forecasts combined with weighted profile that varies across weeks including historical sales from two prior weeks, adjusted for unusual events and product anomalies. In addition, these promotions are accurately timed as per their respective weeks while comprehensive examination is done in order to validate the forecast with respect to past performance. The main goal of this thesis was therefore to assess whether or not the suggested short term forecasting approach was more effective than conventional methods. Through a comprehensive analysis of various parameters and metrics, the comparative study aims to identify the advantages and potential drawbacks of each methodology.