As conceptions of academic development expand to encompass a more diverse and flexible set of supports, traditional approaches to evaluation, including impact studies of formal programming, become insufficient. A program may appear ineffective when evaluation ignores additional supports that interact to counteract implementation or alternatively satisfy individuals' needs. Recognizing the inherently interactive nature of academic development within systemically complex social contexts, this paper examines four evaluative lenses (traditional, ecological framework, complexity theory, and developmental) applied to graduate students' support. By selecting an appropriate evaluation lens, based on local purposes and context, academic developers are better positioned to assess and improve supports provided.