This study aims to develop G-MADE, a generation-responsive design evaluation model based on Multimodal AI, and to verify its validity and applicability. Conventional BX (Brand Experience) design evaluation has mainly relied on qualitative and quantita...
This study aims to develop G-MADE, a generation-responsive design evaluation model based on Multimodal AI, and to verify its validity and applicability. Conventional BX (Brand Experience) design evaluation has mainly relied on qualitative and quantitative assessments by human consumers or expert groups. However, such approaches have inherent
limitations in terms of time, cost, and subjective bias. In particular, establishing design strategies that reflect the diverse values and emotional receptivity of different consumer generations is essential for corporate competitiveness, yet systematic research enabling scientific measurement and comparison has been scarce. Therefore, this study introduces the image–text processing capacity of multimodal AI into BX design evaluation, proposing an objective and reproducible model that incorporates generational responsiveness.
The research first reviewed prior studies to identify the multidimensional elements of BX design, structuring them into three major categories—sensory, emotional, and cognitive/behavioral/relational experiences—and nine detailed evaluation criteria. AI prompts were then designed to reflect generational consumer characteristics. For analysis, major brands with symbolic and representative status in the Korean coffee franchise industry were selected. Visual materials such as logos, store interiors, and advertising videos were collected and evaluated using both multimodal AI and generational consumer surveys. A total of 550 responses were obtained from the surveys, and the same evaluation items and scales were applied to enable direct comparison with AI results.
The analysis revealed a high degree of correlation between AI evaluations and human assessments across multiple items, with similar generational response patterns observed. Generation X highly valued stability and clarity in design, Generation Y responded positively to emotional and aesthetic values as well as storytelling elements, and Generation Z showed distinct preferences for creative and interactive designs. AI evaluations reflected these tendencies, demonstrating its potential to replicate and reproduce generational design responsiveness. However, some items (likeability, emotion inducement, and tactility/materiality) showed weaker correlations with human assessments, indicating the need for refinement. Accordingly, these items were excluded in the final G-MADE model to provide clearer evaluation standards.
Furthermore, an expert FGI (Focus Group Interview) examined the validity and applicability of the AI-based evaluation model. Experts acknowledged the objectivity and efficiency of AI evaluation, while also suggesting the integration of contemporary social values and evolving cultural trends for future expansion. The findings confirmed practical applications in areas
such as BX strategy development, pre-verification of brand experience designs, and generation-tailored campaign planning.
This study quantitatively evaluated generational BX design using multimodal AI and compared the outcomes with human assessments, thereby presenting both academic and practical contributions of AI-based design evaluation frameworks. Academically, it supplements existing qualitative approaches by establishing a repeatable and systematic research framework.
Practically, it offers companies substantial evidence to swiftly and efficiently establish BX design strategies that reflect generational consumer characteristics. Future studies should extend evaluation criteria, compare different AI models, and expand applications across various industries to further enhance the generalizability and reliability of G-MADE. Ultimately, this research demonstrates the feasibility of a convergent approach between AI and human assessment in generational BX design evaluation, presenting a new paradigm for design evaluation studies.
Keywords: Multimodal AI, BX Design, Design Evaluation, Evaluation Model, G-MADE