A Comparative Analysis of Brand Personality of Mobile Network Operators in Thailand and Philippines Using LLM-as-a-Judge: A Case Study of AIS and Globe

บทคัดย่อ

This study explores the brand personalities of leading mobile network operators in Thailand and the Philippines—Advanced Info Service (AIS) and Globe Telecom—using Large Language Models (LLMs) within the LLM-as-a-Judge framework. Three generative AI models, ChatGPT, Gemini, and DeepSeek, were employed to analyze and interpret brand-related content, offering an independent, AI-driven perspective on brand personality. Coca-Cola, a globally recognized consumer brand, was included as a control brand to provide a contrasting baseline outside the telecom sector. The study applied a qualitative content analysis approach using consistent prompts and controlled AI generation conditions to extract brand personality traits. By comparing outputs across multiple AI models, the study highlights how LLMs can consistently identify core traits such as innovative and reliable traits, while also revealing differences in how each model perceives emotional and cultural nuances in brand personality. The findings contribute to marketing and AI research by introducing a novel application of LLM-as-a-Judge for cross-cultural brand personality analysis. This approach expands the use of AI in branding studies beyond technical applications and highlights the role of cultural context in shaping brand personality perception. The results offer practical insights for telecom operators seeking to refine their brand positioning and provide a methodological foundation for future studies exploring brand personality using AI. A major contribution of this study is demonstrating that generative AI models can effectively uncover both shared and unique brand personality traits across industries and regions, offering a scalable approach to cross-cultural brand analysis.