AI Models Outperform Humans in Predicting Purchase Intent, Study Finds
Large language models can now forecast consumer purchase decisions with 90% accuracy, surpassing traditional marketing tools. Researchers from the University of Mannheim and ETH Zürich developed a method called "Semantic Similarity Rating" that converts free-text responses into structured survey data. This breakthrough mimics human test-retest reliability on 9,300 real survey responses.
The technique transforms open-ended AI responses like "I'd definitely buy this" into numerical Likert scale ratings. While promising for market research, it raises critical questions about algorithmic bias and whether synthetic consumers can truly replace human decision-making. The study underscores AI's growing influence in behavioral prediction—a development with potential implications for cryptocurrency marketing strategies and consumer analytics.
