The rise of synthetic respondents in market research: - NIQ

Have you heard? NIQ and GfK are now one website. Learn more

Education

The rise of synthetic respondents in market research:

Why some will make it and some will fake it.

Education

The rise of synthetic respondents in market research:

Why some will make it and some will fake it.


  • Synthetic respondents are artificial personas generated by machine learning models to mimic human responses. When informed by diverse datasets, these “stand-in consumers” can be used to quickly evaluate new product concepts.
  • The overnight rush to launch synthetic feedback tools has posed a dilemma for the market research industry, primarily due to AI’s ability to produce convincing—but sometimes unsubstantiated—output.
  • In this article, we share three characteristics of best-in-class synthetic models—and why a “fake it ‘til you make it” approach won’t suffice.

Understanding synthetic respondents in market research


Imagine: Your team has just leveraged a large language model (LLM) to develop some exciting new snack concepts. But now you’re faced with a new challenge: Which ideas are worth pursuing, which need a few tweaks, and which should be discarded?

For most innovators, the next step is culling the long list of ideas. This is sometimes achieved by using focus groups, but more often by simply selecting the team favorite(s) before moving on to consumer testing—kicking off your innovation cycle with a game of chance. Thankfully, the latest Generative AI (GenAI) technology—which has already demonstrated great promise for product innovators—now has an answer for that, too.

Synthetic respondents are artificial personas generated by machine learning models to mimic human responses in market research. They can represent target markets, specific demographics, or even consumption profiles. Informed by a diverse set of data sources, these “stand-in consumers” can generate feedback on any number of questions, but for product innovation, their most useful application (for now) is to quickly evaluate and optimize new concepts. This early-stage check holds potential to sort through ideas and accelerate innovation cycles while conserving time and resources for research questions that require real human consumer feedback.

But leveraging this technology within an effective solution takes more than attaching a shiny interface to an LLM. The overnight rush of research vendors to release non-optimized solutions with bold claims about their capabilities has led many—ourselves included—to raise an eyebrow. At the heart of the issue is the ability of current LLMs to excel at producing results that seem convincing. Even when these fast-to-market models lack access to the right tools or data for accuracy, they can still generate outputs that pass a gut check. Producing convincing answers is different from providing accurate ones—especially when it comes to making business decisions that rely on data integrity. This discrepancy is also why you might still hesitate to replace your lawyer with ChatGPT.

If you’re intrigued by the promise of GenAI but are evaluating its capabilities with caution, you might be wondering whether synthetic models are a realistic proposition at all. Can a synthetic persona truly convey human consumers’ preferences and opinions? Are there characteristics of best-in-class tools that set some apart from others? How can we ensure these solutions deliver both accuracy and actual business value? With these questions top of mind, we set out to find answers through our own rigorous experimentation.

Building the ultimate synthetic model


Our experimentation in building synthetic models models began years ago, utilizing our in-market transactional data for volumetric forecasting and trend detection. With more than 20 years of experience building patented AI and machine learning solutions—not to mention our vast stores of consumer-panelist data—we’ve always been ideally positioned to bring such a tool to market. But even as the recent emergence of GenAI has propelled new possibilities forward, our experimental approach has remained a deliberate one, prioritizing client value over hype. Above all, we’ve been laser-focused on ensuring output accuracy through testing and validation and are unwilling to compromise quality over speed to market. This strategy has been affirmed time and again, not only through our findings but also in conversations with clients who have noted that the early-market technology they adopted did not deliver as promised.

Best-in-class synthetic models should test, calibrate, and validate response accuracy across every category


Best-in-class synthetic models should leverage the latest granular data to drive accuracy


Best-in-class synthetic models should place data in context to navigate next steps


Navigating the AI revolution together


Synthetic respondents are not a replacement for human consumers in market research; they are a supplement to your ideation process when time is of the essence. As game-changing as their potential might be, these models require unique data sets, prompting, and continuous calibration to consistently generate substantiated and actionable output. Synthetic feedback should be trusted only when the supplier has access to data that validates its accuracy. Thankfully, businesses wishing to explore this emerging technology for market research can avoid future burnout and skepticism by asking key questions of their vendors.

The AI revolution continues to disrupt, with both unprecedented opportunities and significant risks. At NIQ, our commitment to innovation is matched by our dedication to rigorous validation and refinement. We invite you to join us in navigating this transformative era, ensuring that together, we can harness the true power of GenAI for the future of consumer insights. 

Ready to take your innovations to the next level with Generative AI? Contact us to learn about synthetic respondents beta testing at NIQ BASES.

About the authors
About the authors

Martin Levanti is Vice President of Analytics Commercialization at NIQ BASES, where he plays a pivotal role in shaping AI-powered capabilities that enhance research tools and deliver valuable insights to global clients.


Courtenay Verret is Vice President of Global Thought Leadership at NIQ, where she translates consumer intelligence into actionable information through storytelling.