<![CDATA[This study analyzes Pertamina's public relations crisis communication strategy in responding to the issue of “fake fuel” that went viral on social media in 2025. The rapid spread of public complaints, allegations of fuel adulteration, and increasing criticism on platforms such as X and TikTok had a significant impact on the company's reputation, requiring a data-driven approach to crisis communication. Unlike previous studies, which generally relied on manual content reading, this study offers a new approach by integrating machine learning-based sentiment analysis to map public perceptions in a more structured manner. Data was collected by gathering 1,000 posts on the X platform and supplemented with monitoring data from Brand24. All texts were processed using Natural Language Processing (NLP) techniques and classified with a Support Vector Machine (SVM) algorithm verified through ten iterations of Monte Carlo Cross-Validation. This model produced an average accuracy of 0.559 and showed a strong dominance of negative sentiment in 603 posts. Analysis of public engagement on TikTok showed a variety of responses, ranging from support for fuel distribution activities to sharp criticism of service quality and operations at fuel filling stations.These findings indicate that Pertamina's crisis communication strategy has not been entirely successful in reducing negative public perception. Theoretically, this research contributes by integrating machine learning data into SCCT analysis, thereby providing a more accurate understanding of public responses. Practically, the results of this study are expected to help energy companies improve their crisis communication strategies to be more responsive and effective.]]>
Copyrights © 2025
