This study aims to derive effective strategic directions for short-form video advertising by analyzing consumer comments posted on advertising content across major platforms, including YouTube, Instagram, and TikTok. A total of 37,284 English-language comments were collected from diverse sources, encompassing a range of product categories —such as fashion, food, electronics, and entertainment—and various formats of short-form advertisements. The collected data were analyzed using Term Frequency (TF), Term Frequency –Inverse Document Frequency (TF-IDF), BERT-based sentiment analysis, and the NRC emotion lexicon to examine both quantitative and qualitative consumer responses. The results revealed that positively valenced keywords such as “like,” “love,” and “good” ranked highly in both the TF and TF-IDF analyses. Furthermore, positive emotions such as anticipation, joy, and trust were the most prominent, according to sentiment analysis. In contrast, negative keywords such as “soggy” and “bad,” as well as negative emotional reactions—including anger, disgust, and fear—were also observed. These negative responses were associated with ad avoidance, diminished trust in brands or products, and aversion to advertising creative. Based on these findings, this study proposes key short-form advertising strategies, including (1) emotional keyword-centered narrative design, (2) trust-based advertising leveraging consumer recommendations or reviews, (3) expectation-driven serial ad formats, and (4) messaging that emphasizes ethical and safety considerations. By conducting an empirical analysis grounded in consumer discourse, this research provides a multidimensional understanding of cognitive and emotional responses to short-form advertising, thereby offering meaningful theoretical and practical implications for the planning and execution of future digital advertising strategies.