INTEGRATION OF BIG DATA WITH MACHINE LEARNING FOR PREDICTIVE SALES ANALYSIS: A SYSTEMATIC REVIEW

Abstract

This study examines the implementation of Machine Learning techniques in sales forecasting, highlighting the impact of Big Data integration in transforming business strategies. Large-scale companies like Amazon, Google, and Microsoft are leading this shift, using Machine Learning to enhance the accuracy and efficiency of trend forecasting, thereby strengthening their competitiveness. However, small and medium-sized enterprises (SMEs) face significant challenges in adopting these advanced technologies due to limitations in infrastructure and expertise, restricting their ability to leverage predictive sales analysis. Predictive analysis allows for demand forecasting and process improvement, but SMEs encounter obstacles when trying to implement complex techniques such as Random Forest or neural networks, primarily due to data complexity and variable selection. This study evaluates the main Machine Learning techniques used in sales forecasting and the specific challenges SMEs face. The methodology follows PRISMA guidelines and uses the PICO framework to organize the search in the Scopus database. The analysis addresses strategies applied by large corporations and the barriers SMEs face in implementing similar technologies. Additionally, technological solutions that improve sales forecasting efficiency and accuracy are explored, overcoming obstacles such as scalability and data quality. This approach provides a comprehensive overview of the current state of research in sales forecasting, highlighting challenges and proposing strategies for SMEs to adopt Machine Learning and Big Data, helping to reduce the competitive gap with large companies.