Most public K-12 schools in the United States enact a school improvement plan each year based upon the state testing results from the previous year. Despite the many frameworks, research, and recommendations, few scholars tackled what the plans contain and the results. Using a sequential explanatory mixed methods model, two stages examined school improvement plans for language arts tests over three years: a general linear model for 1316 schools in two states and a longitudinal thematic analysis of plans within one district. Most plans produced minor improvements in test scores, and most schools developed generic, nondescript plans to improve teaching practices. The school improvement plans failed to create substantive, visible change in most schools. Instead of the usual recommendations, a simple view of improvement finds schools need to focus on three criteria at the local level: students, curriculum, and highly effective teachers. Each criterion has several subvariables, but school improvement plans should move beyond a focus on student failure and seek to improve all students.