Introduction: Despite studies highlighting the inaccuracies of self-assessment, practicing physicians continue to rely on self-perception to maintain clinical competence. Many approaches have been proposed to augment physician performance. In the realm of Quality Improvement (QI), Audit and Feedback (A&F) has a modest effect. Educators have proposed coaching interventions and academic constructs have invoked training for early-career clinicians. Very few of these are driven by the perceptions and the needs of the end-user - the physicians. We currently lack a model to understand physicians’ perceptions of their own practice data and an understanding of the factors which would enable practice change. In this study, we sought to develop a model for data feedback which may best help physicians change practice. Methods: In a previous study, we conducted a needs analysis of 105 physicians in the Hamilton-Niagara area in order to understand which data metrics were most valuable to physicians. Using the survey results, we designed an interview guide that was used as a qualitative study of physicians’ perspectives on A&F. By intentional sampling, we recruited 15 physicians amongst gender groups, types of practice (academic vs community) and durations of practice. We conducted this interview with all 15 participants which were then transcribed. We then performed thematic analysis and extraction of all interviews using a realist framework. These were then translated into broader themes and, by using a grounded theory framework, created a model to understand how physicians relate practice data to their own sense of self. Interviews were anonymized and no identifying data was shared as part of the interview. All interviewees consented to participation at the outset and could withdraw at any time. Results: Via stakeholder interviews from 15 key informants, we developed a model for the understanding of how a physician's sense of self and the nature of the data (quantity and quality) may be combined to understand the likelihood of practice change and the adoption of the change strategy. Using this model, it is possible to understand the conditions under which A&F would provide the greatest opportunity for practice change. Conclusion: Physician identity intersects with A&F data to shed insights on practice improvement. Understanding the core identity constructs of different physician groups may allow for increased uptake in A&F processes.