Determining Annual and Monthly Sales Targets for Stores and Product Categories in FMCG Retail


Mert B., Eskiocak D. I., Ogul I. U., Aslan M. K., Karalar E.

4th International Conference on Intelligent and Fuzzy Systems (INFUS), Bornova, Turkey, 19 - 21 July 2022, vol.505, pp.524-530 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 505
  • Doi Number: 10.1007/978-3-031-09176-6_60
  • City: Bornova
  • Country: Turkey
  • Page Numbers: pp.524-530
  • Keywords: Retail, Machine learning, Forecasting
  • Yıldız Technical University Affiliated: No

Abstract

The retail industry, especially fast moving consumer goods (FMCG) retail, has many variables affecting sales. Therefore, it is of great importance for companies to accurately plan their annual sales target. The aim of this study is to determine the annual sales target of Migros T.A.S, one of the largest FMCG retailers in Turkey. These targets are then communicated to the marketing teams and stores monthly. This process was previously done manually by the Sales, Marketing, and Finance departments. With this project, both annual and monthly sales forecasts are created for each store and product category using machine learning methods. Within the scope of this study, we create various features, and compare the results of various state-of-the-art algorithms such as Random Forest, LightGBM, and Prophet. By using machine learning we aim to both automate manual processes and determine a baseline sales estimate without bias. Afterward, these base model outputs are optimized by including business information of the relevant departments.