Synthetic respondents are one of the most debated applications of AI in market research, and for good reason. Many off-the-shelf models have struggled to meet the quality and reliability standards researchers expect, leading to widespread skepticism.

As training and validation techniques improve, however, new opportunities are emerging to apply synthetic data in ways that are both scalable and scientifically sound.

In this session, you’ll hear directly from Qualtrics about their latest advancements in synthetic data. You'll get an inside look at how the Qualtrics Edge Audiences model is trained and validated. Learn how we're improving its accuracy, better mirroring human response patterns, and expanding what's possible in synthetic survey research.

You’ll come away with a clearer view of how synthetic respondents can support your research goals, where human input is still essential, and how to evaluate the right use cases for your work.

What You’ll Learn:

• How foundational model design directly impacts the quality, reliability, and usefulness of synthetic research data.
• How to evaluate synthetic data quality and reliability, with a look at how Qualtrics validates its models before using them in research.
• When synthetic respondents work well, when they don’t, and how to know which to use for your project.