Figure 3.
The model shows four labelled columns: antecedents, mediators, moderators, and outcomes. Under antecedents are four boxes labelled generative A I capability, L L M driven emotional adaptability, conversational naturalness and human likeness, and algorithmic transparency and explainability. Arrows from these boxes point to two mediator boxes labelled personalisation fit and co-creation quality. From personalisation fit, arrows lead to emotional engagement. From co-creation quality, arrows lead to personalisation fatigue slash cognitive overload. Emotional engagement and personalisation fatigue slash cognitive overload, both point to guest satisfaction. Guest satisfaction points downward to brand loyalty. A separate arrow from personalisation fit points to perceived experience quality, which also points to guest satisfaction. Above these paths are two moderator boxes labelled trust propensity and privacy concern, each connected by dashed arrows to the relationship between personalisation fit and perceived experience quality. A dashed vertical boundary includes a box labelled cultural orientation at the bottom, indicating moderation across paths leading to guest satisfaction and brand loyalty.Conceptual framework of GenAI-driven personalization dynamics
Source: Developed by authors