Customer-Centric Sales Forecasting Model: RFM-ARIMA Approach

Authors

  • Sanket Londhe School of Computer Engineering and Technology, MIT-World Peace University, Pune, Maharashtra
  • Sushila Palwe School of Computer Engineering and Technology, MIT-World Peace University, Pune, Maharashtra

Keywords:

Business Intelligence, Customer Analysis, Sales Forecasting, Exploratory Analysis, Segmentation, Decision Support System, Recency, Frequency & Monetary Value (RFM), Auto-Regressive Integrated Moving Averages (ARIMA), Long short-term memory (LSTM)

Abstract

Background: Decision makers use the process of determining the best course of action by processing, analysing & interpreting the data to gain insights, known as Business Intelligence. Some decision support systems use sales figures to predict future expansion, but few consider the effect of customer data. Objectives: The main objective of this study is to build a model that will give a forecast based on fine-tuned sales numbers using some customer-centric features. Methods/Approach: We first use the RFM model to segment the customers into distinct segments based on customer buying characteristics and then discard the segments that are irrelevant to the business. Then we use the ARIMA model to do the sales forecasting for the remainder of the data. Results: Using this model, we were able to achieve a better fitment of the data for the prediction model and achieved a better accuracy when used after RFM analysis. Conclusions: We tried to merge two different concepts to do a cross-functional analysis for better decision-making. We were able to present the RFM-ARIMA model as a better metric or approach to fine-tune the sales analysis.

Downloads

Published

2022-09-28

Issue

Section

Information Systems Research Articles