Intelligent Customer Segmentation Considering Beer Sales Based on Beer Attributes, Products and Price: A Case Study for Districts of Istanbul

Abstract

This study aims to perform intelligent customer segmentation considering beer sales based on beer attributes, products, and prices for 39 districts of Istanbul. A case study is carried out taking into account the beer sales of a beer producer in Istanbul for 2018, 2019, and the first 9 months of 2020. In this regard, k-means clustering process, which involves the hierarchical centroid linkage clustering algorithm, is employed to group predetermined five clusters, which are diamond, gold, silver, bronze upper, and bronze lower, based on six variables (liter, price, mouthfeel, bitterness, alcohol ratio and aromaticity). Values of Recency, Frequency, Monetary (RFM) are computed for eight products within 39 districts of Istanbul. Furthermore, seasonality analysis is conducted to reveal the effects of coronavirus on beer sales in Istanbul. This study can act as a guideline to predict the future sales for each district of Istanbul considering the features of the product to be released.