A CMO’s Take on Esomar North America 2026: From Better Insights to Greater Impact
At the AI panel at 2026 ESOMAR North America in Atlanta, Ed Keller shared a number that surprised me. Over the previous 90 days, four of the ten most-viewed videos at Market Research Institute International (MRII) were about AI. The single most-viewed video was about how to write a great questionnaire.
That made me smile. Everyone is talking about connected knowledge bases, synthetic personas and digital twins, and they should. Still, people want to learn how to ask good questions. I agree with them. The way we ask and the way we engage people decide what evidence we end up with.
I attended many sessions, and two themes kept coming up. Better participant experiences helped researchers learn more. And teams that worked closely with the business were able to act on what they learned.
The first one is close to home. In 2018, the founders of Rival Technologies and Reach3 Insights set out to change the research experience. People were being asked to share their thoughts and feelings in surveys that felt long and impersonal, nothing like how they talk in daily life. We believed a better participant experience would bring richer, more thoughtful answers. In Atlanta, I saw that belief reflected in many different projects.
In his presentation, Matt Kleinschmit, our Founder and CEO at Reach3 (a Rival Group company), talked about why conversational research is better by design. The centerpiece of his session is a research-on-research study Reach3 ran comparing traditional online surveys with mobile-first, conversational surveys built via the Rival platform.
Average open-ended responses went from 10 words in a traditional survey to 25 in mobile chat. With AI probing, they reached 48. Video responses produced nearly eight times the content of traditional open-ended answers. Participants (across all age groups, including Boomers and Gen Xers) also said the experience was more engaging and said they were more willing to take part again.
More words do not always mean more insight. What I look for is detail that helps me understand why someone answered the way they did. A follow-up question gives a person room to explain. Video feedback can show context and emotion that a short written answer leaves out.
Matt then showed what happened when AI analyzed these richer answers. Using Rival’s Unstructured Data Agent, the team reported higher thoughtfulness scores for outputs based on conversational inputs. Built by Rival, Thoughtfulness Score uses AI to measure open-ends across 10 research-grade dimensions. The lesson is simple. Give AI richer answers and it has more to work with.
You can read the research and get the full report on the Reach3 Insights website.
Two more sessions were about context. Chick-fil-A tested digital menu boards with visual attention, simulated ordering and experience feedback. The layout that got the most attention did not always lead to the best simulated ordering results. A preference score alone could not answer the business question.
Unilever let participants use a prototype at home before asking them to explain its benefits. After using it, people had the words to describe something unfamiliar. The team also found a clearer direction for innovation.
In both of these examples, the teams gave people something relevant to respond to. Researchers learned more, and the business knew better what to do next.
Matt’s Mars example brought both themes together. Its grocery research combined conversational surveys, in-store missions and interview debriefs. Then the findings went to stakeholders as retailer-specific materials, video showreels and workshops, so they could use them in customer discussions. The design covered how people contributed and how the business received the findings.
A Pernod Ricard launch study was my favorite example of acting early. Researchers saw that people knew about the campaign, but few had tried the product. The team shared this during fieldwork and moved budget toward tastings. The activation plan changed while there was still time to act.
This is what I want from research as a CMO. I make decisions about customers, positioning and investment without complete information. I want to know early who will use the findings, what they can change and when they need to know. Those answers should shape the work from the first brief.
These themes matter even more as research becomes connected. Model Context Protocol (MCP) connections can give AI access to research tools and data. Knowledge bases let teams build on what they already know. Synthetic personas and digital twins give us new ways to explore questions. All of this depends on the evidence underneath it, how current it is and where we still need fresh human input.
At Reach3, a sister company of Rival Group, we believe that evidence must also come from people who are real and who they say they are. A rich answer is only useful if you trust where it came from. That is why we bring together conversational technology, access to people and research expertise.
I left Atlanta with two questions I want us to keep asking. Are we creating the conditions for people to tell us something meaningful? And are we helping someone make a better decision with it?