AI–Human Hybrids for Marketing Research: Leveraging Large Language Models (LLMs) as Collaborators.

This article investigates the integration of large language models (LLMs) as collaborators in the marketing research process, demonstrating that a human–LLM hybrid approach enhances efficiency and effectiveness in both qualitative and quantitative research. Empirical studies conducted with a Fortune 500 food company show that LLMs can generate synthetic respondents, moderate in-depth interviews, and analyze unstructured qualitative data with performance comparable to or exceeding human-only methods, particularly in depth and insightfulness. For quantitative surveys, LLMs replicate response patterns well, and incorporating context through few-shot learning and retrieval-augmented generation (RAG) improves the heterogeneity and internal consistency of synthetic data. The authors propose practical road maps for adopting LLMs in marketing research while emphasizing the necessity of human oversight to mitigate biases and ensure data quality.