AI Models Can Predict Purchases Better Than Humans
A new study reveals that AI models can accurately forecast purchase intent, outperforming traditional marketing tools.
Researchers at the University of Mannheim and ETH Zürich found that large language models can replicate human purchase intent by transforming free-form text into structured survey data.
The "How likely are you to buy this?" metric can be replicated by AI models with striking accuracy.
Methodology
The team introduced a method called "Semantic Similarity Rating," which converts open-ended responses into numerical "Likert" ratings, a five-point scale used in traditional consumer research.
This method allows AI models to mirror human purchase intent, blurring the line between survey respondent and simulator.
Implications
The study suggests that AI models can predict what you'll buy better than you can, making them a valuable tool for marketers.
Author's summary: AI models outperform humans in predicting purchases.