Abstract

Nowadays Social Media platforms play an essential role in both interpersonal communication and business marketing strategies. Hence, identifying and mapping user behavior on these platforms is important. This study explores the impact of different social media post types, including videos, photos, statuses, and links, on user engagement performance metrics such as reactions, comments, and shares. The research focuses on data collected and processed from fashion and cosmetics business pages. The extracted dataset consists of 7000 instances, and 12 attributes. The study makes use of data mining techniques, specifically decision trees classifiers, to classify the performance of these post types using the WEKA software. The study managed to provide insights which indicate a significant preference for live videos and photos over other social media post types. The decision tree classifier provides valuable insights about the types of posts that trigger user engagement allowing marketers to extract their social media strategies sets of rules. The findings have a major effect on businesses that aim to optimize their online identity, awareness and reach using targeted and effective social media content. However, due to the focus on a small-medium sized dataset in fashion and cosmetics industries, this study faces limitations. Future research will generalize these findings by addressing these limitations including a wider range of industries and data. Finally, this research contributes to the field of digital marketing by offering valuable insights into online consumer behavior on social media, providing a methodology that can assist businesses comprehend customer needs and predict user engagement.

EL Classifications: C63, D81, D83, L81, L86.