future state
- Research & Data Quality
- Article
Why Synthetic Respondents Flatten Consumer Sentiment
A new ARF Psych of GenAI experiment reveals that large language models apply a rigid, rule-driven logic when evaluating privacy scenarios—even when humans typically shift their reasoning based on framing, emotion and social context. Unlike consumers, who blend intuition, feeling and social perspective into their judgments, GPT-4o relied on a single internal rule across all testing conditions: data use is acceptable only with explicit consent. This consistency offers value for certain analytic tasks but exposes limits for advertising research that depends on emotional nuance and context-sensitive consumer insight.
Member Only Access
- Research & Data Quality
- Article
When AI Takes the Survey: Evaluating LLMs as a New Tool for Consumer Insight
The study in this MSI working paper evaluates whether large language models (LLMs) can serve as a reliable source of consumer preference data—potentially transforming how market research is conducted. Using conjoint-style survey questions, the researchers compared LLM-generated choices with human responses to estimate willingness-to-pay (WTP) for a variety of product attributes. They find that LLMs often approximate human preferences surprisingly well, especially when fine-tuned with prior survey data, though important limitations remain. For marketers, the research highlights both the promise and the boundaries of using AI-generated insights to accelerate testing, concept screening and early-stage innovation work
Member Only Access