Abstract

Customer segmentation is the classification of customers based on their attributes that are specific and similar. By using the clustering method, the decision of which customer segment to target is determined, which has been many retailers challenge, it provides an excellent opportunity for internal customer analysis. The dataset used for this analysis has over 3 Lakhs transaction of customers, it gets challenging to derive at the decision as to which group of customers to target to Upsell or Cross sell any product. Hence that is where a company loses most of its marketing investment as there is a high marketing expenditure that goes into the campaigning of any sales, but not many sales are successful and there is a revenue to the company from it. Hence low Return on Investment (ROI). This is the problem that can be resolved by Customer Segmentation. The research focuses on segmenting clients utilizing customer-specific data such as Customer IDs and over all consumption levels and it is achieved by using machine learning methods – Specifically used in this research are, the Density Based Clustering Algorithm (DBSCAN) and the Partitioning Algorithm (KMeans Clustering Algorithm). The findings indicate that there are 3 distinct groups of customers or the campaign target, each with unique purchasing characteristics on whom the company’s campaigning marketing design can be based on.

Keywords: Customer Segmentation, Cluster, DBSCAN, Kmeans

Conference Name: World Conference on Artificial Intelligence: Advances and Applications (WCAIAA 2023)