AI promises speed and efficiency, but without the right foundations, CMOs risk faster outputs with shallower insights. To truly gain competitive advantage, the real ROI lies in the learning to be able to see the full picture.
AI promises speed, scale and efficiency. For many marketers, that’s reason enough to dive in. But Thomas Kellermann, head of insights at GUT Network, argues that if you’re only focused on faster outputs, you’re missing the point.
The real risk, he says, isn’t just generic work or homogenized outputs, it’s the loss of learning that comes from working with the data itself.
“Efficiency gains might get you to output faster. But if you skip the process, you don’t learn. You lose touch with the intelligence – the full context. And then you can’t see the whole picture,” he says. “The models are what they are. The difference is whether the foundation is right. That’s where the learning happens.”
Advertisement
Marketers have rushed to experiment with generative AI for content and creative development, where 62% of CMOs say they are seeing the biggest impact right now. But Kellermann believes too many are thinking about it only in terms of cost savings or speed to market.
“The perception today is that AI is a way to do the same thing at a cheaper price, or faster. But that’s not good enough. We need to take all those benefits and actually create something better,” he says.
That requires starting at the input side and not just optimizing for outputs.
AI models will only ever be as good as the foundations beneath them. For Kellermann, that means asking harder questions about how data is structured, enriched and interpreted before it ever hits an algorithm. “You can’t just throw everything at it and expect it to make sense. You need the right scaffolding, the right foundation,” he says. “Without that, the outputs are just noise.”
The challenge right now isn’t lack of data but a lack of harmony. Brands need to structure and enrich data by domain and discipline, so that when AI is applied, it’s drawing from a coherent, contextual foundation.
His message to CMOs: don’t hand too much over to AI. “I’m not a fan of outsourcing everything. A lot of learning comes from doing – from spending time with the data. If you hand it all over, you lose that understanding,” he says.
That’s not just a philosophical risk but a practical one. Without the learning that comes from direct engagement with data, CMOs risk losing their ability to validate, contextualize and ultimately trust what AI delivers.
Advertisement
Fixing these foundations isn’t just a job for data scientists. It requires leadership from the very top of marketing. “It has to be solved by everybody. But the CMO is the one who can set the directive – to make sure there’s a unified view of the company’s data. That becomes your interface for informing the AI systems,” he says.
In other words: the CMO must champion a frictionless, structured approach to insights – ensuring that data is not just available, but usable, and being used to elevate the understanding of the customer. AI can amplify insights. But only if CMOs invest in the discipline, scaffolding and frameworks that make those insights meaningful.
The future of marketing intelligence won’t be about who moves fastest – it’ll be about who learns best.
Explore more in the AI Marketing Pulse hub.
