Mind the Gap: How LLMs Identify, Update, and Represent Brands After Repositioning
(A2026-134060)

AUTHORS

Andreas Hamann, University of Mannheim; P. K. Kannan, University of Maryland; Florian Stahl, University of Mannheim

ABSTRACT

While large language models (LLMs) are increasingly emerging as new touchpoints in the customer journey and shaping brand visibility, it remains unclear how their representation of brands changes when a brand repositions, a strategic move frequently undertaken by firms. In this study, we build on brand positioning theories to examine how LLMs identify, update, and represent brand information after repositionings. By conducting a synthetic field experiment, we obtain three key findings. First, LLMs vary systematically in how fast and how accurately they identify and reflect repositionings. Second, repositionings temporarily reduce brand mentions in competitive brand queries. Third, firms can influence how their brands are represented and ranked through search engine re-indexation and repositioning communication practices. Our findings extend concepts on brand meaning formation to generative AI-mediated contexts and provide actionable insights for managing LLM-based brand visibility.