What industry leaders said at InsightsOn 2026
Yale InsightsOn 2026 last Thursday. Five sessions. 75+ senior insights and marketing leaders. One main theme running throughout the day: the insights function isn't the problem. The system built around it is.
Here's what we heard – and why it matters.
Insights aren't the problem. The system around them is.
Maharaj Singh, SVP Strategy North America at The Coca-Cola Company, said it best: "If insight does not change a decision, it is just expensive decoration."
The diagnosis was consistent across the conference. Amanda Fraga, SVP of Global Marketing Strategy, Insights & Transformation at Mastercard, put it plainly: "Marketing needs to move in real time. Current systems haven't kept pace – decision-making is often still fragmented, manual and hard to scale." Ryan Verschoor, Global VP of Insights & Capabilities at AB InBev, was equally direct: "AI allows for one source of truth so we're all singing from the same song sheet. Nothing is worse than people talking about different numbers in front of leadership – it drives down the impact of the insights function.
The problem isn't that insights teams aren't doing good work. It's that the infrastructure connecting that work to the people who need to act on it is broken. Insights get produced. Decisions get made. The two rarely meet at the right moment.
AI owns the what. Humans own the so what.
This was the reframe that ran through almost every session, and it matters because it rejects both the panic and the complacency that tends to surround AI conversations in marketing.
Stan Sthanunathan, Executive Chairman of i-Genie.ai, put it cleanly: "In a world of abundant answers, deep human insights becomes the scarce asset.” The division of labor they described is clear: AI owns problem definition, pattern recognition, and signal mapping at scale. Humans own human truth, the judgment to act, and the ability to tell a story that moves an organization.
Colleen DeCourcy, CMO of Sonos, pushed this even further: "We went from labor to cognition. That's the job we have currently. And the next level is meta-cognition – thinking about the thinking. This is an explosive area for humanity, and I can't think of a place where we have to do more thinking about our thinking than insights." When AI handles the generation, the human's job shifts to evaluation: deciding what's worth acting on, what's missing, what the data can't see. That's not a smaller job. It's a harder one. The insights leaders in that room aren't being replaced. They're being promoted – but only if they have the courage to show up with a point of view.
Marie Gulin-Merle, GVP of Ads & Commerce at Google, added something that should be on the wall of every insights team: "AI helps reduce the distance between me and the actual signal." That distance – the layers of process, translation, and time between a real human truth and the person who needs to act on it – is where good decisions go awry.
The winners are building systems, not running projects.
If the first two themes diagnosed the problem, this one is the prescription. And the leaders in that room weren't talking about it abstractly. They were describing what it looks like in practice.
Stan Sthanunathan again: "If the old world was about projects, the new world is about systems." Dr. Elaine Rodrigo, Chief Insights & Analytics Officer at Reckitt, showed what that looks like at scale – a pipeline that compressed the full insight-to-validated-concept process from 12 weeks to 6 hours, with 80-100% adoption across 1,200 marketers in 11 months. Reckitt built this internally, assembling best-in-class partners to make it work. The results are undeniable: 4x faster concept generation, +24% effectiveness on their See My Pain campaign vs. non-AI concepts, and 100% of AI-assisted concepts passed with a Superior profile.
AB InBev's Ryan Verschoor described a roadmap to move from data provider to strategic growth engine by 2028. The plan has three stages: build the technical foundation, strengthen internal capability, and by 2028, operate as the company's strategic growth engine – not a support function. Mastercard also built an internal system – theirs combines business intelligence, real-time cultural signals, and simulated personas. Instead of replacing human judgment, the system puts better inputs in front of it faster.
The through-line: every insights leader who is winning right now has stopped thinking about research as a series of projects and started building the connective tissue between signal and decision. Always-on. Continuous. Integrated into the moment where choices actually get made.
The Bottom Line
The message across every session was the same: the insights function is more important than ever, but the way it operates has to change. Projects become stale. Decks get filed away. The organizations pulling ahead aren't doing more research – they're building systems that connect real human signal to decisions at speed.
That's exactly what Suzy's new Decision Engine does. From human signal to deep insight to presentation-ready assets, in minutes. So your team spends less time on the what, and more time on the so what — which, as everyone in that room made clear, is where the most important and impactful work happens.
