This thesis explores the effectiveness of various sentiment analysis techniques on the TikTok social media platform, particularly in the context of fashion-related content. Given the popularity of TikTok among Generation Z, understanding sentiment on this platform is crucial for marketers and businesses targeting this demographic. The study employs four distinct sentiment analysis methods: Support Vector Machines (SVM), VADER (Valence Aware Dictionary for sEn- timent Reasoning), Convolutional Neural Networks (CNN), and Bi-directional Long Short-Term Memory networks (biLSTMs), to analyze user-generated con- tent on TikTok. Data were collected through scraping TikTok videos related to fashion brands, followed by manual labeling to prepare for sentiment analysis. The analysis revealed varied accuracy performances across the methods, with each showing strengths and weaknesses depending on the specific characteris- tics of the data. The results indicate that no single method outperforms others consistently across all types of data, highlighting the complexity of applying SA to the informal and evolving content on TikTok. This research contributes to the understanding of sentiment analysis applications in new social media con- texts and suggests directions for future work, including expanding the amount of data analyzed, the range of content and platform features included, and the adaptability of sentiment analysis tools to changes in online language use. The insights gained are particularly relevant for developing more nuanced sentiment analysis strategies aimed at researching the digitally native Generation Z on emerging social media platforms.