Data-driven product marketing strategies: An in-depth analysis of machine learning applications

Abstract

Marketing is a planned method of creating, communicating, delivering, and exchanging valuable solutions for customers, clients, partners, and society as a whole. Machine learning techniques may completely dominate the market based on big data analysis, whereas traditional measuring tools fail to capture the market value of big data. Using a high-level overview and implementation process, this article explains how ML is being used in marketing. The purpose of this paper is to deliver the findings of an investigation into the connection between marketing strategy and the creation of successful new products over the long term. This study presents an in-depth analysis of marketing product strategies through an application of ML models, specifically CNN-LSTM, DT, and SVM. Utilizing the Amazon Product Review dataset, the research involves extensive data preprocessing, including steps like tokenization and stop word removal, followed by the splitting of data. The models' performance is evaluated using metrics including F1-score, recall, accuracy, and precision. The CNN-LSTM outperformed the others with a 94% accuracy rate, while Logistic Regression achieved the lowest at 70%. The research highlights the value of sophisticated machine learning models for improving marketing tactics, leading to better marketing-related decision-making.