Colour and social media engagement in fashion images: an exploratory cross-cultural image mining study in the U.S. and Indonesia

Instagram has become a prominent platform for communication and commerce, particularly in the fashion sector. This study explores the impact of colour features (warm hue, saturation, brightness, and contrast of brightness) on Instagram users’ engagement with fashion images in two countries, the U.S. and Indonesia. Using visual analytics tools such as OpenCV, we analyzed 4,537 Instagram posts from the U.S. and 3,320 from Indonesia. The results revealed a cross-cultural difference: images with warmer hues garnered higher engagement in the U.S., but fewer likes in Indonesia. Saturation and brightness negatively affected engagement in both countries, whereas contrast of brightness had no significant effect. The findings contribute to the research on colour and cross-cultural marketing and offer valuable practical insights for global fashion brands. Based on the image mining results and prior literature, we also offer propositions to guide future research.