Competing for Attention: Text Overlays and Showing Faces in Nonprofit Instagram Posts

Nonprofit organizations increasingly rely on visually rich social media posts to capture attention and mobilize support. Among emerging design practices, text overlays (e.g., words or phrases deliberately inserted within an image) are gaining prominence, yet their interaction with other salient visual cues remains underexplored. This study investigates how text overlays influence engagement and whether this effect depends on the simultaneous presence of human faces depicted in the image. To address this gap, we analyzed 1982 Instagram posts published between April 2023 and June 2025 by UNICEF International, Save the Children US, and Doctors Without Borders. Images were annotated through a computational content analysis approach powered by generative artificial intelligence, which simultaneously detected text overlays and face presence. The findings reveal that text overlays significantly increase engagement, but this positive effect disappears when human faces are present, indicating that the impact of textual cues depends on the surrounding visual context. These results advance understanding of multimodal nonprofit communication and provide practical guidance for designing more effective social media appeals.