Empirical studies in teaching effectiveness research (TER) have largely yielded small and inconsistent effects. To better understand and explain these effects, we argue that it is core to consider how teaching and its quality have been conceptualized and defined in TER. Looking both within and beyond TER, we identify and discuss three potential contributors: lack of a common definition of teaching and its quality and overemphasis on effects on student academic achievement in the definition of teaching quality, a simplistic assumption that there is a universally valid conceptualization of teaching, and inaccurate conceptualization of the dynamics among different aspects of teaching. To address these challenges, we offer three steps related to giving conceptualization a more prominent role in TER; defining and using terms more clearly, consistently, and transparently; and enriching, extending, or revising TER models to better embrace the complexity of the work of teaching.