The Most Appropriate Avatar Design for Sri Lankan Online Consumer Banking: A Comparison of 3D vs. 2D Designs

Abstract

AI Avatars are graphical representations of human beings, and the modern world tends to use avatars in their online web applications in order to enhance customer satisfaction. This study aims to investigate the optimal avatar model for the Sri Lankan online customer banking sector by examining whether 2D or 3D avatars perform better. The study has taken Generation Z customers as the sample. Using predefined avatar characteristics, such as Likeability (LI), Experience (EX), Credibility (CR), Attractiveness (AT), Pleasure Emotions (PE), Arousal Emotions (AE), Dominance Emotions (DE), Buying Intention (BI), and eWOM (WM), the study evaluates the effectiveness of each model. The analysis is conducted through the SPSS and Mean-value-Based Decision-Making methodology with an evaluation of nine hypotheses. Based on a representative sample of Generation Z, the research findings reveal that 2D avatar models outperform their 3D counterparts across most attributes, indicating their greater suitability for enhancing user experience and fostering customer engagement in Sri Lanka's online banking platforms. The findings will fulfil a novel concept by contributing to the knowledge gap in existing literature.