When Generative AI Persuades Better: Cognitive Constraints, Personalization, and Promotional Effectiveness

The rapid integration of generative AI (GenAI) into commercial platforms offers unprecedented opportunities for scalable, personalized promotional communication. This study examines the effects of GenAI-based production promotion on customer purchase decisions in the context of online food ordering. Partnering with a chain restaurant, we conducted a randomized field experiment in which customers were assigned to receive one of four types of promotional messages: AI-generated personalized messages, AI-generated generic messages, human-generated generic messages, or coupons only. Results show that both AI-generated message types significantly enhance customer purchasing behavior compared with the coupon-only group, with AI-personalized messages increasing daily order amounts by 14.3% and AI-generic messages by 10.4%, while human-generated generic messages show no significant improvement over the control. Moderation analyses reveal that AI-generic messages are particularly effective in high cognitive load contexts (e.g., in the workplace or during working hours), while AI-personalized messages maintain effectiveness across all contexts. We employ the Elaboration Likelihood Model (ELM) to identify a two-tier cognitive mechanism underlying these findings. First, AI-generated messages are more readable and present higher-quality arguments than human-generated messages, reducing comprehension barriers and enriching the substance available for persuasion. Second, we find that AI-generated generic messages leverage strong peripheral persuasion cues. In contrast, AI-generated personalized messages incorporate both peripheral and central route elements, enabling dual-pathway persuasion. This research contributes to the Information Systems (IS) literature by providing causal field evidence on the promotional effectiveness of GenAI, advancing cognitive load and ELM theories in the context of AI-generated content, and offering practical insights for AI-driven customer engagement strategies.