This study explores how phygital profiling and dynamic personas enhance personalized customer experiences across the evolving customer journey. Drawing on Foursquare data from 27,148 users and over 2 million check-ins across four global cities, we construct digital twins using location-based social network graphs. We compare the effectiveness of phygital versus digital profiling and dynamic versus static personas in two key tasks: point-of-interest (POI) recommendation and social connection mapping. Results show that phygital profiling significantly outperforms digital profiling, and dynamic personas yield more accurate, real-time personalization than static approaches. These findings underscore the value of integrating physical and digital data in capturing the complexity of modern consumer behavior. The study offers theoretical and practical insights for advancing adaptive personalization and customer engagement strategies in increasingly hybrid consumption environments.