Monitoring behaviour to influence behaviour

A friend of mine asked me today (actually it was my alter ego but journalistically quite convenient): “why is it that you are so obsessed with tracking, monitoring and observing behavior in /KL7?” “I’m very glad you asked” I replied. And this is what I said:

Look at this picture. Why is this woman crying? (get the answer below…)

Trying to understand peoples behavior and design ways to influence it by means of traditional analytics is flawed. You cannot trust what people answer in questionnaires let alone base your strategy on it. Here’s a telling example: In a survey we conducted for a client in the health sector more people knew which values the brand represented than the brand itself! It is hardly necessary to inform you that this is logically impossible! They simply guessed (based on prejudice. More on that below). And when asked what would make them consider a cheaper foreign competitor they answered ‘quality’. Admitting price-sensitivity when it comes to your own health is taboo in Denmark so respondents found a clever way out. But the truth value of the answer was very low and not basis for strategic choice.

With a term borrowed from neurology we could call this phenomena cognitive ‘filling in’: When meaning or knowledge is not available we simply fill in the meaning or knowledge ourselves. Examples in cognitive science are legio. E.g. in visual or semantic experiments. If you are shown incomplete symbols, sentences or other holistic cognitive entities the brain fills in the missing parts with you not even noticing it. In the examples above a certain level of animosity was at play, since the client’s sector is  not fully covered by public welfare and the client’s members thus perceived as ‘greedy’ given the fiscal relation very foreign to danes in a health context.

The point is that cognition is holistic and make the data/parts/answers fits the result/overall perception. That is why you need to study people as they really are rather than ask them how they think they are is order to engineer their behavior. But why is there a picture of a crying woman in the post you might ask? Because we do the same narratively. When asked why the woman is crying most people answer: Because her loved one just left her. And they will be likely to act like this ‘fact’ was knowledge – not simply a guess.