Aims: Patient experience surveys are often used for purposes of quality improvement and governance of person-centred practices. Reporting results of patient surveys typically foregrounds the average experience. Consequently, experiences of some members within diverse populations may be rendered invisible or misrepresented. Personas (hypothetical representation of patients) are a tool to help amplify hidden voices by relating responses to patient experience measures to patients’ life stories. A mixed-methods approach to creating data-driven personas (Figure 1) was used to better understand and attend to unique primary care communication experiences identified through latent class analysis. Methods: A latent variable mixture model (LVMM) was applied to examine heterogeneity of responses to six communication-related questions from 3,539 people across Alberta. Measurement bias was taken as the difference in factor scores between a model accounting for heterogeneity and a model not accounting for heterogeneity. Sixteen diverse respondents belonging to the class with the greatest measurement bias were invited to participate in an interview, five of whom identified as Indigenous. A constellatory approach was used for the preliminary mapping of six unique personas. Acknowledging that Indigenous experiences are often less attended to, interviews with Indigenous patients were foregrounded as the core constellation of four personas.  Results: The LVMM resulted in improved fit for a 2-class compared to a 1-class model. Class 2 consisted of 24% of survey respondents who responded to communication questions more negatively, relative to class 1. The latent class information and interview experiences informed the creation of six interactive visual personas. Figure 2 provides an initial visual of Jesse (persona 1) to better understand experiences of chronic pain management in rural primary care.   Conclusions: Personas can be used as practice-based scenarios to support healthcare practitioners in applying person-centred care principles. This can ultimately tailor care to individual needs that have historically received less attention.