Useful information
Prime News delivers timely, accurate news and insights on global events, politics, business, and technology
a new research work Released quietly last week, it describes an innovative method that allows large language models (LLMs) to simulate human consumer behavior with surprising accuracy, a development that could reshape the multibillion-dollar business. market research industry. The technique promises to create armies of synthetic consumers that can provide not only realistic product ratings, but also the qualitative reasoning behind them, at a scale and speed currently unattainable.
For years, companies have tried to use AI for market research, but have been hampered by a fundamental flaw: When asked to provide a numerical rating on a scale of 1 to 5, LLMs produce unrealistic and poorly distributed responses. A new newspaper, "LLMs reproduce human purchase intention by obtaining semantic similarities from Likert ratings," submitted to the arXiv preprint server on October 9, proposes an elegant solution that avoids this problem entirely.
The international team of researchers, led by Benjamin F. Maier, developed a method they call semantic similarity rating (SSR). Instead of asking an LLM for a number, SSR asks the model for a rich textual opinion about a product. This text is then converted into a numerical vector: a "embed" – and their similarity is measured against a set of predefined reference statements. For example, a response from "I would definitely buy this, it’s exactly what I’m looking for." would be semantically closer to the reference declaration for a "5" qualification than to the declaration of a "1."
The results are surprising. Tested on a huge real-world data set from a leading personal care corporation (comprising 57 product surveys and 9,300 human responses), the SSR method achieved 90% test-retest reliability in humans. Crucially, the distribution of ratings generated by the AI was statistically almost indistinguishable from that of the human panel. The authors state, "This framework enables scalable simulations of consumer research while preserving the metrics and interpretability of traditional surveys."
A timely solution as AI threatens survey integrity
This advancement comes at a critical time, as the integrity of traditional online survey panels is increasingly threatened by AI. A 2024 analysis of the Stanford Graduate School of Business highlighted a growing problem of human surveyors using chatbots to generate their responses. These AI-generated responses were found to be "suspiciously nice," too detailed and lacking "sarcasm" and authenticity of genuine human feedback, leading to what the researchers called a "homogenization" of data that could mask serious problems such as discrimination or product defects.
Maier’s research offers an entirely different approach: Instead of fighting to remove tainted data, it creates a controlled environment to generate high-fidelity synthetic data from scratch.
"What we are seeing is a shift from defense to offense," said an analyst not affiliated with the study. "The Stanford paper showed the chaos of uncontrolled AI contaminating human data sets. This new article shows the order and usefulness of controlled AI when creating your own data sets. For a data manager, this is the difference between cleaning up a contaminated well and tapping into a fresh spring."
From text to intention: the technical leap behind the synthetic consumer
The technical validity of the new method depends on the quality of the text embeddings, a concept explored in a 2022 paper in EPJ Data Science. That research advocated for a rigorous "construct validity" framework to ensure that text embeddings (the numerical representations of the text) actually "measure what they are supposed to do."
The success of the SSR method suggests that their additions effectively capture the nuances of purchase intent. For this new technique to be widely adopted, companies will need to be confident that the underlying models not only generate plausible text, but map that text to scores in a robust and meaningful way.
The approach also represents a significant leap from previous research, which has largely focused on using text embeddings to analyze and predict existing online review ratings. TO study 2022For example, it evaluated the performance of models like BERT and word2vec in predicting review scores on retail sites and found that newer models like BERT performed better for general use. The new research goes beyond the analysis of existing data and generates novel and predictive insights before a product even hits the market.
The dawn of the digital focus group
For technical decision makers, the implications are profound. The ability to spin a "digital twin" of a target consumer segment and testing product concepts, advertising copy or packaging variations in a matter of hours could dramatically accelerate innovation cycles.
As the article notes, these synthetic respondents also provide "rich qualitative comments explaining your ratings," offering a treasure trove of data for product development that is both scalable and interpretable. While the era of human-only focus groups is far from over, this research provides the most compelling evidence yet that their synthetic counterparts are ready for business.
But the business case goes beyond speed and scale. Consider the economics: A traditional survey panel for a national product launch could cost tens of thousands of dollars and take weeks to implement. An SSR-based simulation could deliver comparable insights in a fraction of the time, at a fraction of the cost, and with the ability to instantly iterate based on findings. For companies in fast-moving consumer goods categories – where the window between concept and shelf can determine market leadership – this speed advantage could be decisive.
Of course, there are caveats. The method was validated in personal care products; its performance in complex B2B purchasing decisions, luxury goods or culturally specific products remains untested. And while the article demonstrates that RSS can replicate aggregate human behavior, it is not intended to predict individual consumer choices. The technique works at the population level, not the person level, a distinction that is very important for applications like personalized marketing.
However, even with these limitations, the research marks a milestone. While the era of human-only focus groups is far from over, this paper provides the most compelling evidence yet that their synthetic counterparts are ready for business. The question is no longer whether AI can simulate consumer sentiment, but whether companies can move fast enough to capitalize on it before their competitors.
