How Does Brand-Generated Content Engage Consumers? A Study on Fashion Brands' Instagram Data.

This article investigates how design features of brand-generated content (BGC) on Instagram influence user engagement, measured by the number of likes, focusing on 20 major fashion brands divided equally between luxury and mass-market segments. Using machine learning techniques to extract image and caption features and applying Poisson regression models, the study finds that including URLs generally reduces engagement, while multiple images per post increase likes. Distinct preferences emerge between brand categories: luxury brand audiences favor posts emphasizing products with minimal distractions and selective use of emojis and mentions, whereas mass-market brand audiences respond positively to posts featuring mega-influencers, conspicuous backgrounds, and fashion collocation topics. The research offers practical recommendations for optimizing Instagram marketing strategies tailored to brand positioning and highlights the importance of caption text alongside visual content in driving consumer engagement.