Communication flow and feedback loop concept illustrated with wooden blocks

How We’re Using Generative AI in Market Research to Uncover Deeper Human Insights

Michael Patterson, Radius Insights, Radius Global Market Research 2021/10/michael-patterson-bio.jpg

by Michael Patterson, PhD

Chief Research Officer

Recently, our team presented a case study at a Marketing Research Society (MRS) conference focused on AI. They outlined how we leveraged generative AI to uncover deep insights on consumer behavior.

The case study is a good example of how we are using AI to augment our human-centered research with rich insights from data sources such as text, videos, social media, and images to provide our clients with stronger insights. The ability to seamlessly blend qualitative and quantitative methodologies is already improving the way we conduct research and deliver insights to clients. Two areas where we’re seeing the most impact are survey design and respondent quality.

Intelligent Probing transforms the survey experience to gain deeper insights.

We are currently developing an approach to surveys called Intelligent Probing which leverages conversational AI to design a dialogue with survey respondents. The technique provides respondents with personalized prompts that encourage them to share more descriptive answers to open-ended questions.

This approach is an add-on to online surveys. It allows us to gain greater detail and richer insights from open-ended questions. For example, within a quantitative survey we might ask 10–15 minutes’ worth of closed-ended questions, and then tack on a qualitative module where we pose a question almost like a discussion guide. We take the response provided and send it through an AI engine which then sends back a follow-up question designed specifically to prompt the respondent to expand on their answer. If needed, we could send that response back to AI for an additional follow-up question.

The back-and-forth makes the survey feel more intimate to the respondent and encourages them to share on an emotional level. Over the course of a survey, we might add these qualitative modules in a few areas. We’re finding that Intelligent Probing drives deeper engagement. Respondents react positively to the prompts, so we get deeper answers and far fewer nonsense answers.

For example, a conversation might look like this:

Survey question:
How did this ad make you feel?

It made me feel sad because it reminded me of my dad who passed away last year.

AI-based follow-up question:
I’m sorry to hear that. Would you like to share what specifically in the ad reminded you of your dad?

My dad used to take my brothers and me ice fishing at a cabin just like the one in the commercial. Watching the ad made me think of the great time we had, the freezing cold lake and the smell of the warm fire at night. I really loved that experience, and it made me sad in a good way to think about how much I miss him.


This solution is an extraordinary way to integrate quantitative and qualitative approaches that result in deep insights. The level of detail helps brand teams digest the research faster. With specific and emotional feedback from consumers or clients, teams can develop action plans across the organization to connect with customers at an emotional level at the right moments and drive stronger outcomes for their brand.


Serene woman looking at her laptop screen at home

Our insights will always center around consumers and the rich input they provide to us as AI creates new opportunities to explore, test, and deliver insights that help our clients build stronger action plans and growth strategies.”


New AI techniques improve respondent quality and data validation.

Generative AI now plays a key role in helping us detect survey fraud and enhance data quality. These additional tools in our toolbox speed the process of screening respondents to weed out fraudulent survey takers such as people faking credentials to participate in high-paying surveys. We have also improved our ability to spot multiple IP addresses as well as identify and remove bots, in some cases we can do this in real time. (Read more about our commitment to identifying survey fraud.)

Machine learning and AI techniques enhance our data collection processes, increase error detection, and provide speed and accuracy for tasks such as identifying and removing duplicate records. Each small improvement to the way we collect and analyze data increases our ability to deliver accurate insights in a shorter period of time.

Human-centric insights remain at the center of our research approach.

With generative and conversational AI, we can cast a wider net to deliver expanded services to our clients. We can find and mine rich data sources and analyze data with greater speed and confidence. We’re updating our predictive models to boost accuracy, and we’re enhancing the way we build segments and targets. Our insights will always center around consumers and the rich input they provide to us as AI creates new opportunities to explore, test, and deliver insights that help our clients build stronger action plans and growth strategies.


Learn more about how generative and conversational AI can build deeper connections with your customers.

Contact Us