Where the Data Falls Silent

COCenk Onur Ozan ·20 Oct 2026Published 20 Oct 2026 ·4 min read

For years, business has had the same favourite sentence: “We make data-driven decisions.” Lately it has been updated: “We make AI-driven decisions.”

We really are at an impressive point. Millions of data points can be processed in minutes; we know which product a customer looked at, where they gave up, when they bought and which campaign they clicked. Analyses that used to take weeks now fit into a few minutes. To be honest, the amount of data we have has not changed much. The information we once got excited about as big data, which later turned into a kind of data dump because we could not process it, is something we can now dare to work with thanks to AI.

Still, an important gap remains. Seeing behaviour and understanding it are not the same. AI can offer an explanation for why a customer abandoned their basket; it can surface relationships between hundreds of variables such as price sensitivity, delivery time, alternative products and past purchases. But why did the same customer buy a more expensive product from another brand the next day? The model can produce a probability, but it cannot know the real reason. What was on the customer’s mind at that moment may have been a memory stirred by an ad, a friend’s recommendation or simply that day’s mood. This is where the data falls silent, because people’s important decisions are not purely rational.

Kahneman showed years ago that much of our decision-making comes from fast, intuitive and automatic mental processes. Antonio Damasio showed that a mind cut off from emotion cannot decide well; he documented patients whose emotional responses had weakened after brain damage and who, despite intact reasoning, got stuck on even the most ordinary everyday decisions. Gerald Zaltman argued that around ninety-five per cent of our thinking happens outside conscious awareness; the figure has been much debated since, but the direction it points to has not changed. People feel first and then find a reason to justify the feeling. The algorithm usually captures the reason, not the feeling.

Today’s algorithms naturally analyse visible behaviour for the most part: what we do, when we do it and how much. Why we do it usually lives inside the person rather than inside the data, and that is where what we call insight is found.

An experience is less the event itself than the trace it leaves in the mind. People often remember brands for how they made them feel more than for their products. Two customers can receive exactly the same service, call the same contact centre and hear the same answer; one becomes attached to the brand while the other decides never to come back. The difference lies not in the service but in the mood in which it was received. The behaviour is the same, the experience is different. Data notices this difference only later, perhaps as a complaint or a churn rate; people feel it in the moment.

This is also where loyalty starts to be misunderstood. Companies feel the need to reduce it to numbers to make it measurable, which is an understandable reflex. We measure loyalty with repeat purchase rates, explain it with NPS and look for it in CRM records. Yet loyalty is less the purchase itself than the result of the mental and emotional bond that produces it. A customer can buy from the same brand three times, not out of loyalty but because it is the nearest store, because they could not find an alternative, because they caught a discount or simply out of habit. Data may label all of these as loyalty; psychology may disagree.

The opposite is possible too. A customer may not buy for years and still keep positive feelings about a brand, and when circumstances change, that brand is their first choice again. They may happily recommend a brand they used years ago to a friend, even with no interaction today. In the CRM, this customer looks passive or lost; in their mind, the brand is still alive. No algorithm can measure this on its own today, because human behaviour does not move in a straight line; it shifts with context, is shaped by emotion and often contradicts itself.

That is why experience management cannot be the business of data science alone. We need to bring behavioural science, psychology, neuroscience and sociology to the same table. AI is one of the strongest players at that table, perhaps the fastest, but not the only one. Organisations that leave it on its own will try to grow while measuring perfectly what their customers do and never understanding why.

I do not think bigger data sets or more powerful models will be what separates companies in the coming years. The difference will come from organisations that can understand the invisible motivations behind behaviour. Data tells us what the customer does; insight tells us why. The distance between them is sometimes as short as a single question and sometimes as long as years of research. AI will keep reading behaviour, and people will still be the ones who make meaning of it. In your organisation, who closes that distance?

Cenk Onur Ozan COO, Cloud4Feed

Writes about customer experience, behaviour and the decisions that follow from listening.

LinkedIn