Descriptive Analytics of Instagram Content from Indonesian Skincare Brands: Integrating Content Analysis and TF-IDF

Abstract

Instagram has become an important channel for skincare brands to communicate with consumers and strengthen brand positioning. However, limited access to structured data, such as APIs and internal platform insights, requires a systematic approach to using public data. This study aims to map dominant keyword themes, identify message strategy characteristics, and describe engagement variation in the Instagram content of three Indonesian skincare brands—Skintific, Azarine, and Somethinc—during June 2023. The study employs a descriptive analytics approach with a mixed-method mini-case design that combines content analysis and TF-IDF-based text mining. The data were collected from screenshots of public Instagram posts, including captions, hashtags, likes, and comments. The findings show that the communication of the three brands centers on skin concerns, product categories, and benefit and ingredient cues, with a predominantly problem–solution framing. Although they share similar general themes, each brand constructs a different communication emphasis: Skintific through scientific narratives, Azarine through light and accessible educational communication, and Somethinc through experiential and participatory activation approaches. In terms of engagement, comments appear to be more sensitive to interactive content and strong calls to action, whereas likes more strongly reflect initial interest in specific content or campaigns. This study contributes to integrating text analytics and content analysis in the examination of social media content based on public data, thereby supporting content strategy evaluation and digital marketing decision-making. The findings are descriptive in nature and are not intended to explain causal relationships.