An Agent-Based Simulation Model Integrating Generative AI and Fast and Frugal Heuristics for Enhanced Marketing Automation

This paper introduces an agent-based simulation model integrating generative AI with fast and frugal heuristics to enhance marketing automation. The model adopts fast and frugal heuristics as the primary decision mechanism, invoking generative AI only when situations are ambiguous or exceed the scope of rule-based evaluation. Unlike traditional black-box systems, the architecture preserves interpretability by structuring decisions through transparent heuristic gates before any AI-based reasoning. Central to the framework is a Multi-Agent System (MAS) that simulates customer behaviors and marketing strategies through specialized agents. By leveraging a Vector Database (VecDB) of historical data, the model enables effective personalization without customer-specific fine-tuning. The framework, designed as both a research and a strategic decision-support tool, addresses current limitations in automation by adapting to real-time market dynamics while ensuring transparency through an Explainability Insights Agent. Simulation results demonstrate significant improvements in customer engagement and retention, providing a pathway for interpretable AI-human collaboration in marketing.