Unlike general entrepreneurs, streamers generate revenue by creating and delivering self-produced digital content to their audience. Streamers are digital contentpreneurs. Motivated by our observations that streamers offer self-designed visual designs based on their live streaming content, we aim to investigate the effect of these visual designs across different live streaming categories and how they ultimately influence streamers’ entrepreneurial performance during their live sessions. We employ the Stimulus-Organism-Response (SOR) Model and the Person-Environment (PE) Fit theory as baseline models to develop our hypotheses. To test these hypotheses, we will use large language models and deep learning methods to extract visual design details from streamers’ past live streaming videos.