Phytopharmaceutical plants have become one of the main commodities contributing significantly to the economy through their use in the pharmaceutical, cosmetic, and health industries. However, behind this economic potential, traditional herbal medicine businesses often face challenges, particularly in promotion and brand identity. Social media platforms like Instagram have now introduced unique features to support business and marketing, primarily by providing in-depth information about herbal products and offering opportunities for businesses to receive feedback from consumers. Comments on social media are valuable but often unstructured; hence, sentiment analysis is necessary to organize and categorize this data. By combining comment data with information from Google Trends, cause-and-effect relationships from comments during specific periods can be identified using path analysis. This research aims to analyze consumer comments on the Sidomuncul company's Instagram platform, with the hope of benefiting the company and advancing herbal medicine products. The methods used in this study include Artificial Neural Network (ANN) and K-nearest neighbor (KNN) to classify comments into positive, negative, and neutral categories. Both methods show satisfactory results in classification, with an average accuracy of 0.887 for ANN and 0.874 for KNN. However, the ROC curve for the KNN model indicates a relatively low AUC value in classifying negative comments, at 0.598.