Customer behavior is an area of study in retail marketing and retail decision making. Various factors of human nature and sociology affect the purchasing behavior of the customer. For companies to meet these different needs and desires, it is of great importance to better define and understand their customer groups. This study aims to predict the next purchase day and accordingly clarify brand campaign strategies and apply appropriate communication tools in line with customer needs by identifying the customer groups to be prioritized. In this way, the aim is to improve customer satisfaction by increasing interaction through intelligent foresight. For this purpose K-Means has been used in Recency-Frequency-Monetary (RFM) analysis and Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), Multi-layer Perceptron (MLP) and Naive Bayes (NB) have been used for day prediction. The performance of the developed models has been evaluated using accuracy and F1-score. The results show that it possible to predict the next purchase day with reasonable error rates using single-classification and multi-classification. With this study, it is expected that the planning activities of the brand will be facilitated in terms of communication and an increase in conversion rates will be experienced with the right marketing budget investment.
Anahtar Kelimeler: Recency-Frequency-Monetary Analysis, Machine Learning, Purchase Day
|
