Firms worldwide increasingly use competency-based performance appraisals, and a growing proportion link competence proficiency appraisals to rewards. This “best practice”, highly disseminated by consultants and other HR pundits, is seem as a key tool to align human capital to business strategy and performance. Its main assumption is that competence mastery at the individual level would drive up unit- or organizational-level performance. Thus rewarding proficiency in individual competencies would be justified by it boosting organizational results. However, few empirical studies tested the empirical validity of this assumption. We conducted a large-scale inquiry involving “live” and consequential competence appraisal data for 682 managers in a large bank. We tested if appraised proficiency in six soft skills competencies, linked to rewards, influenced unit- level performance. Our study indicated that unit-level performance could not be significantly explained from individual-level competency, but insist that further research is needed. Based on these findings, we argue that organizations presenting similar disassociation between individual-level appraised competencies and unit-level performance should consider decoupling their behavior-based performance appraisal from their compensation and reward decisions.