Design Science Research (DSR) begins with problem identification, yet limited guidance exists on what constitutes a “good” research problem, and how to find it. Problems articulated in the business environment are often too specific to yield generalizable contributions. Conversely, grand challenges that DSR could contribute to are often too abstract to be addressed rigorously in a single project. We address this gap with a conceptual framework for specifying and evaluating DSR research problems. Drawing on both design science and design thinking methodology, we develop two complementary models: (1) a Problem Funnel showing how problems decompose from grand challenges to research-tractable problems, and (2) a Problem Space Matrix positioning problems along two independent dimensions: scope and structure. (3) We propose seven criteria for evaluating problem statement quality: appropriate specificity, empirical grounding, stakeholder significance, tractable solvability, generalizable scope, solution novelty, and articulable clarity, and (4) develop five guidelines for problem identification. The paper develops a workflow consisting of the Problem Funnel, Problem Space Matrix, criteria, and guidelines, which we illustrate with examples from sustainable behavior and enterprise systems research. By doing so, we contribute to DSR methodology by providing systematic guidance for the central, yet surprisingly insufficiently understood activity of problem formulation.