A metric is often presented as a neutral window onto an existing category: a school is 'effective,' a neighborhood is 'safe,' or a worker is 'productive,' and the number merely reveals how much. But selecting what to count can reshape the category itself. If school effectiveness is measured only by test gains, instruction and resources may reorganize around tested outcomes; unmeasured aims become less visible. Metrics do not invent reality from nothing, yet they partly constitute the practical meaning of what institutions claim to measure.