Technical Debt (TD) decisions are rarely just technical. Stakeholders consider heterogeneous criteria, such as deadlines, client expectations, architectural consequences, available resources, organizational authority, and team sustainability, when deciding whether to acquire, postpone, prioritize, or repay TD. However, these criteria are often reported under different labels and at different levels of abstraction, making it difficult to compare findings across TD decision contexts. This paper proposes a practitioner-informed taxonomy represented as a conceptual model of stakeholder criteria for TD decision-making. Based on a qualitative study with 11 software practitioners in Brazil, we extracted, consolidated, and organized criteria related to TD acquisition and repayment. The resulting taxonomy comprises six families: stakeholder-facing value; delivery and resource pressure; technical integrity and systemic risk; decision basis and epistemic style; governance and legitimation; and human and team sustainability. Our findings show that acquisition and repayment share taxonomic families but differ in decision function. In acquisition, criteria often operate as permission mechanisms that make shortcuts acceptable under current constraints. In repayment, criteria operate as authorization mechanisms that justify allocating time and resources to address existing debt. The conceptual model further shows how criteria become actionable through stakeholder interpretation, translation across technical and business perspectives, and organizational legitimation. The taxonomy and model provide an empirically grounded artifact for classifying TD decision criteria, comparing decision contexts, and structuring discussions about TD acquisition and repayment.

翻译：暂无翻译