Agencies Tap Into Benefits of AI-Powered Focus Groups

Planners can ask sensitive questions, cut costs and time, drive campaign success

Mark your calendar for Mediaweek, October 29-30 in New York City. We’ll unpack the biggest shifts shaping the future of media—from tv to retail media to tech—and how marketers can prep to stay ahead. Register with early-bird rates before sale ends!

Brands and their agencies are increasingly using artificial intelligence-generated focus groups to simulate conversations with potential customers in order to inform campaigns, saving time and money and asking more sensitive questions.

At ad agency Supernatural, chief strategy officer Mike Barrett asked an AI-generated persona what the most impactful challenges of cancer are, informing the campaign for a healthcare brand.

“It’s not the treatment or chemotherapy,” the virtual cancer patient responded. “It’s the fact that nobody knows how to talk to you about it. It feels lonely and isolating.” This insight led to the agency creating 12 AI-generated personas for the brand.

“You would never ask a [real] person that question,” Barrett told ADWEEK.  

Generating audience data from market research can often take several weeks and cost thousands of dollars. Three agencies—including Pereira O’Dell, Supernatural and Media.Monks—implementing AI-generated personas for campaigns, using brands’ first-party data to inform cohorts, found that AI-generated focus groups are a cost-effective solution yielding results within days.

They also don’t deviate too far from human focus groups: One study from December 2023 found that data produced by large language models aligns with the data garnered by human participants between 75% and 90% of the time.

“Generative AI’s ability to stand in for a focus group is emerging as one of its most impactful use cases for marketers,” said Gartner vice president distinguished analyst Andrew Frank. “In addition to delivering more testing and brainstorming capacity at lower cost, it also addresses some of the privacy issues limiting the availability use of real data.”

More data, site traffic and conversion rates

Supernatural developed AI models for various patient categories across women’s health and fertility for the healthcare client.

“We generally see results from [AI focus groups] that are better [compared with real focus group] because we have way more data,” said Barrett.

A second healthcare brand that used AI focus groups to inform campaign insights saw a 250% increase in site traffic, leading to a 25% jump in conversion rates.

Media.Monks recently unveiled Persona.Flow, an AI-generated audience tool driven by the Retrieval-Augmented Generation framework. According to Michael Dobell, the agency’s chief innovation officer, this approach provides LLMs with real-time access to brand data, effectively reducing the risk of hallucination. Nearly 10 brands within the agency are testing Persona.Flow.

The agency used Persona.Flow to inform the strategy for a product launch pitch for a financial institution brand targeting Generation Z. Insights from three virtual cohorts noted that this younger group is trend-conscious, inspired by novelty and the thrill of discovery, and prefers brands with a social conscience. The AI focus groups noted that this group is made up of budget-minded shoppers who seek products that serve a functional purpose and foster a sense of identity and community.

Using Persona.Flow reduced the time spent on audience insights by 15%, the agency told ADWEEK.

“Virtual customers would have far more statistical significance than just a one-time focus group,” said Dobell. “When the dataset is pulled from household data across millions of Americans, you’re getting a much more statistically significant response.”

Enjoying Adweek's Content? Register for More Access!