Medical school instructors face a challenge in covering high volumes of material with a large group of students, in a short period of time. Unfortunately, this often means that not all concepts will be sufficiently communicated, understood, and recalled. Internet videos (e.g., Khan Academy), have made it easier for educators to curate course content online, and while the prospect of simply shifting course material out of live instructional time towards recorded online formats is tempting, it is a zero-sum game—medical students still will grapple with overwhelming volumes of information and challenging concepts. Instead, we propose that instructors make use of existing student interactions during office hours to efficiently collect information about student mental models and then use and inexpensive presentation-recording software to create short, targeted videos that disseminate recommended alterations to the presented student mental models. We developed this algorithm in a first-year medical physiology course—which can be highly conceptual and require students to go through multiple iterations of conceptual models. 92.5% of students responded “the videos enhanced their learning in physiology. Nine-three percent of students said the videos helped explain a difficult concept. However, analysis of exam scores suggested that the videos did not improve performance. Based on this feedback, we believe that our algorithm does have a positive impact on student learning experiences, and may have an unmeasured effect on student performance. In sharing this work, we hope to encourage other medical educators to document, analyze and evaluate their pedagogical experiences, frame them in theory, and publish them in an academic format. We believe that the majority of medical educators currently teaching students have excellent training in other domains evaluation and analysis and hope to see such talents pointed towards the field of medical education.