AI Models Accurately Predict Consumer Purchase Intent
- A study by the University of Mannheim and ETH Zürich shows large language models (LLMs) can predict purchase intent with up to 90% reliability.
- The research used a method called “Semantic Similarity Rating” to convert free-text responses into structured survey data.
- The method was tested on over 9,300 human survey responses about personal-care products, achieving results that closely mirrored real consumer data.
This study highlights the potential for LLMs to replace traditional marketing tools by accurately predicting consumer behavior at a lower cost and with greater speed. The findings could transform how companies conduct product testing and market research.
Despite its success in this domain, the method’s generalizability to other product categories remains uncertain, raising questions about its broader applicability and ethical implications. Source
