Comparative Analysis of Machine Learning Algorithms for Enhancing Social Media Marketing and Decision-Making in Kenyan SMEs

Small and medium-sized enterprises (SMEs) in Kenya are crucial to the nation's economic advancement, yet they sometimes have difficulties competing in a rapidly digitalizing market due to limited resources and inadequate marketing strategies. Social media platforms such as Facebook, Instagram, and X (formerly Twitter) are essential tools for cost-effective marketing; nevertheless, many SMEs fail to leverage their potential due to a lack of data-driven strategy. Machine Learning (ML) algorithms offer a transformative method for SMEs to examine social media data, enhance campaigns, and refine decision-making. This research conducts a comparative analysis of five prominent machine learning algorithms: Logistic Regression, Decision Trees, Random Forests, Support Vector Machines (SVM), and Neural Networks, with the objective of improving social media marketing campaigns and decision-making for SMEs in Kenya. The researchers assess the effectiveness of these algorithms in critical marketing functions, including consumer segmentation, sentiment analysis, and campaign optimization. A dataset comprising engagement indicators, customer profiles, and campaign performance metrics from Kenyan SMEs was used to evaluate the algorithms' accuracy, precision, recall, F1 score, and computational efficiency. The findings demonstrate that Random Forests strike a balance between accuracy and computational efficiency, making it a feasible choice for small and medium-sized enterprises with constrained resources. Logistic Regression is cost-effective and suitable for basic jobs, while Neural Networks are proficient at handling unstructured data but require significant computer resources. Decision trees, despite being understandable and user-friendly, are prone to overfitting, whereas support vector machines, although effective for small datasets, require significant computational resources for large-scale applications. The research indicates that significant challenges, such as insufficient technical expertise, elevated computing expenses, and data privacy issues, hinder the use of machine learning by small and medium-sized enterprises in Kenya. It also highlights the potential of cloud-based machine learning platforms, support from the government and private sectors for SME training, and partnerships to improve the accessibility of machine learning solutions. This research contributes to the growing body of knowledge on the application of ML in marketing and provides actionable recommendations for Kenyan SMEs to harness ML technologies for improved social media marketing and informed decision-making.