Revision is the process of improving the quality of translations. While several frameworks for assessing the products of translations/revisions, such as revision parameters and translation issue categories, are available, metalanguages for describing the effects of individual revisions have not been well established. Here, we have developed a metalanguage that can be used by revisers and other players involved in translation projects. In this chapter, we first briefly examine the notion of revision effects and define the basic components of metalanguage, namely, quality attributes and their aspects. We also present our metalanguage, developed by data-driven procedures, to systematically analyse the actual revision instances. Our metalanguage currently covers nine quality attributes, such as “accurate” and “adapted,” with their aspects, such as “meaning,” “terminology,” and “domain convention,” which enable fine-grained descriptions of how and in which aspect a revision improves translation.