Developing a KPI Framework for Intermediate Stage Evaluation in External Research

Sammanfattning

Intermediate-stage R&D project evaluation represents a critical yet understudied governance challenge in complex, multi-domain portfolio environments. At the decision point between fuzzy front-end exploration and resource-intensive feasibility engineering, organizations must assess project success potential using preliminary information, under high uncertainty, and across heterogeneous project types, conditions that render informal judgment and generic checklists both inconsistent and indefensible. This thesis addresses two interconnected research questions: which Key Performance Indicators are particularly relevant and operationally feasible for evaluating intermediate-stage R&D project success potential in automotive R&D, and how these KPIs should be operationalized into a consolidated framework enabling transparent, consistent, and explainable intermediate-stage decisions. The study was conducted at Volvo Cars, a context characterized by a portfolio of over 200 externally co-funded R&D projects competing for constrained resources across safety, electrification, connectivity, and manufacturing domains. A qualitative single-case research design combined document analysis of five internal governance artifacts with thirteen semi-structured practitioner interviews, triangulated through a four-rule procedure, retention-with-refinement, consolidation, addition, and drop, to translate literature-derived constructs into an empirically validated framework. A systematic literature review across six streams, project success theory, ex-ante evaluation methodology, Multi-Criteria Decision Analysis, regulatory complexity, intellectual property strategy, and organizational knowledge management, derived ten candidate KPIs meeting three selection criteria: ex-ante assessability, objective measurability, and non-triviality. Empirical testing confirmed all ten as decision-relevant and operationally feasible, with one additional construct, Sustainability Impact Score, emerging from practitioner interviews and integrated through the triangulation procedure. Tier 1 (Mandatory) contains Strategic Alignment Confirmation, Regulatory Pathway and Risk Index, and Organizational Capability Match as foundational prerequisite considerations that are expected to be addressed before scored evaluation proceeds. Tier 2 (Primary) contains Technology Readiness Level, Risk-Adjusted Value Estimate, Risk-Adjusted Business Potential, and Minimum Viable Evidence Standard as weighted scoring criteria that structure and guide the stage evaluation discussion toward an informed resource-commitment decision. Tier 3 (Secondary) contains Competitive Differentiation Mapping, Real Options Value, IP Strategic Positioning, and Sustainability Impact Score as contextual qualifiers informing deliberation without entering the composite score. The tiers interact through explicit conditioning logic, ensuring that disqualification, differentiation, and contextual calibration operate at explicitly separated logical levels, preserving practitioner deliberation as the final arbitration mechanism. At the practical level, the framework offers Volvo a structured basis for intermediate-stage guidance. It contributes to theory by operationalizing Regulatory Pathway and Risk Index and IP Strategic Positioning without established intermediate-stage protocols, demonstrating MCDA-based qualitative framework development through a replicable four-rule triangulation procedure. Limitations include single-organization scope, cross-sectional evidence, and the absence of longitudinal outcome tracking; directions for future research include predictive validity testing, multi organization comparative studies, and standalone construct development for inter-organizational R&D governance.