In this article we highlight analyses conducted in two qualitative literacy studies to discuss various implications of a blended, or hybrid, approach to multimodal analysis. By investigating several prominent frameworks commonly used together for the purpose of analyzing multimodal data, and describing our own experiences blending these frameworks, we determine that a hybrid approach is not necessarily ineffective at producing data interpretations, but that it is insufficiently reflexive of the role researcher positionality plays in multimodal analysis. We conclude the article by offering recommendations for supplementing hybrid analytical approaches through data co-construction and increased attention to researcher positionality.