Regardless the importance of counting on software requirements metrics, few indicators are used in practice to support the requirements management during software projects. The main reason seems to be the effort required to obtain these metrics, which usually involve manual calculations. Based on a literature review, this paper identifies twenty-two metrics that were reported as useful for requirements management, and establishes the level of automation that can be reached to calculate each metric. Based on it, the paper presents a classification that helps project leaders identify particular indicators to deal with the typical problem sources affecting the requirements engineering processes. Adopting appropriate metrics allows software teams and project leaders to improve the success rate of software developments.