Analogies have an ambivalent status in archaeology. On the one hand, analogies with known societies can suggest rich and compelling interpretations of the past. On the other hand, analogies seemingly rely on the unfounded assumption that the past resembles the familiar. Previous methodological discussions of this challenge have often focused on the adequacy criteria for one specific use of analogies: (i) providing evidential support for interpretations. This paper identifies two further uses: (ii) generating new potential interpretations, and (iii) providing reasons for pursuing an interpretation, i.e. reasons for testing and developing it further. A systematic philosophical analysis of all three uses is presented and it is argued that each of them are subject to different adequacy criteria. Thus, methodological discussions of analogies in archaeology should avoid conflating these three uses. The fruitfulness of this framework is illustrated through a case study involving the interpretation of Pompeian household artefacts.