Useful generalisations, useless specifics.

In our early history, survival depended on quickly weighing risks and reacting. So we evolved the ability to attach values to things. Red means danger. Darkness is risky. Unfamiliar faces call for caution. A smile is usually good. Food that smells bad is repulsive. A sudden loud noise is a bad sign.

This habit of snap judgement spread because it worked. Guesses based on generalisations were more accurate than random ones, and humans became very good at informed guesswork. That's why prejudice is a natural instinct that operates below conscious awareness. Only by recognising it can we understand it and, when necessary, resist it.

What we're taught as children and what we experience build these associations. They shape how we feel about big concepts like race, religion, gender and nationality, and about trivial ones like cars, ketchup and fashion labels. They're learnt, and riddled with errors. But they've served us well for a very long time, so we can't simply ignore them.

Take an example. On average, older people are less comfortable with new technology than younger people. As a rule of thumb, that's broadly true. But assume any particular 70-year-old can't use a smartphone, and you'll often be wrong, and probably patronising. Our built-in risk radar makes useful generalisations, but they lead to useless specifics.

This matters for anyone in marketing. We spend our days building generalisations: audience models, segments, profiles, creative strategies. At the general level, they make marketing more efficient. But they should never be applied to the specific, and too often they are. Businesses start to believe their individual customers are their customer profiles. They aren't. They're unique individuals.

The conclusion is simple but profound. Generalisations based on group characteristics are useful for assessing general risk, improving efficiency and making strategic decisions. As a tool for judging an individual, they're inaccurate, risky and often just wrong.