Proofs of Concept (PoCs) are used in software-intensive systems to evaluate technologies, reduce uncertainty, and support early decision-making. Despite their role in architectural practice, PoCs remain conceptually underdefined and methodologically ad hoc in both research and practice. In this position paper, we argue that PoCs should be a first-class architectural decision instrument rather than informal experiments or disposable prototypes. We highlight the conceptual ambiguity surrounding PoCs and discuss how this ambiguity undermines architectural decision-making. We then articulate a structured, architecture-centric perspective on PoCs as short-lived, risk-driven experiments explicitly designed to inform architectural choices. Drawing on insights synthesized from academic and grey literature and illustrated through industrial scenarios, we outline a lightweight framework that positions PoCs as repeatable, analyzable, and decision-oriented artifacts. We conclude by surfacing open challenges intended to provoke discussion on how architectural decisions are currently made in practice.

@inproceedings{paperbot11645,
  title = {Proofs of Concept as First-Class Architectural Decision Instruments},
  author = {Fabio Petrillo and Bruno Antognolli},
  booktitle = {ICSE},
  year = {2026},
  doi = {10.1145/3774748.3795875}
}