<![CDATA[Song popularity on digital platforms has increasingly influenced public music preferences. However, differences in popularity metrics across Spotify, YouTube, and TikTok have resulted in fragmented approaches to identifying popular songs, preventing a comprehensive representation of song popularity. This study aims to identify song popularity characteristics by integrating cross-platform popularity metrics using the K-Means clustering algorithm. A quantitative approach was employed using primary data collected from questionnaires administered to 71 respondents and secondary data consisting of 500 songs selected from the Most Streamed Spotify Songs 2024 dataset. The data were analyzed using Z-transformation normalization and the K-Means clustering algorithm. The findings identified three principal popularity characteristics—Low Popular, Viral Popular, and Stable Popular—demonstrating that TikTok virality does not necessarily reflect overall song popularity across digital platforms. This study contributes by proposing a conceptual framework for Popular Song Identification based on the integration of cross-platform popularity metrics, offering a more comprehensive approach to identifying popular songs for society, the digital music industry, and the development of music recommendation systems. Abstrak Popularitas lagu pada platform digital semakin memengaruhi preferensi musik masyarakat. Namun, perbedaan metrik popularitas pada Spotify, YouTube, dan TikTok menyebabkan identifikasi lagu populer masih dilakukan secara parsial sehingga belum mampu merepresentasikan popularitas lagu secara menyeluruh. Penelitian ini bertujuan mengidentifikasi karakteristik popularitas lagu melalui integrasi metrik lintas platform menggunakan algoritma K-Means Clustering. Penelitian menggunakan pendekatan kuantitatif dengan data primer berupa kuesioner yang melibatkan 71 responden dan data sekunder berupa 500 lagu yang dipilih dari dataset Most Streamed Spotify Songs 2024. Data dianalisis menggunakan normalisasi Z-Transformation dan algoritma K-Means Clustering. Hasil penelitian mengidentifikasi tiga karakteristik utama popularitas, yaitu Low Popular, Viral Popular, dan Stable Popular, serta membuktikan bahwa viralitas TikTok tidak selalu merepresentasikan popularitas lagu secara menyeluruh. Penelitian ini berkontribusi melalui pengembangan kerangka konseptual Popular Song Identification berbasis integrasi metrik popularitas lintas platform sehingga menawarkan pendekatan yang lebih komprehensif dalam mengidentifikasi lagu populer bagi masyarakat, industri musik digital, dan pengembangan sistem rekomendasi musik.]]>
