Online shops fail to predict customer behaviour

The Political Editor of Sky News, Adam Boulton, is now a YouTube sensation after his live TV spat with Labour Party adviser, Alistair Campbell. “Don’t keep saying what I think,” yelled Mr Bolton, “don’t keep telling me what I think.” There is no love lost between these two anyway; but this row exposes more than rivalry between political enemies. At its heart is a fundamental piece of psychology – we just hate it when other people tell us what we think or predict our behaviour in any way.

Indeed, according to new research if we detect that anyone is trying to predict our thoughts, we do our damnedest to fool them. Even if the prediction of our behaviour is true, we work hard to prove it isn’t. So, how does that leave the likes of Amazon? This company is often held up as a shining example of how to conduct business online because of its varied techniques to get us to part with cash – often based on predicting what we might like.

Through its “recommendation” system, Amazon uses your searches and purchases to try and predict what else you might be interested in. However, according to research about to be published by the Society for Personality and Social Psychology this approach is flawed. Psychologists at the University of Michigan have demonstrated that people often go out of their way to avoid being predicted or to fool the company trying to make the prediction. In fact, according to the researchers, there appears to be a deep-seated response which makes us try our hardest to be unpredictable if we detect someone is trying to predict our behaviour.

Interestingly, the study found that we do not mind being predicted if we perceive the company doing the prediction as some kind of partner. When we feel they are co-operating with us as an individual, then we do not mind being predicted. In other words, if you are obviously – rather like Amazon trying to boost your sales – then most people will work hard against you in attempting to do this. If, however, you are perceived as being someone who is trying to help, in a sense of co-operation, then predictions do work.

Think of it like this. Consider your behaviour when one of those interminably long Internet sales letters gets your cash and then – just before you are to pay – it pops up a “special offer – exclusively for you”. It has predicted you will “want” this item. Sometimes it works; mostly it doesn’t. Indeed, frequently it causes the lack of the original sale. Now compare that with a ” personal shopper” in a large department store. When this person wanders around with you, helping you “choose” clothing for a special occasion they are clearly predicting what you want to buy – but we know they are working for the shop in a bid to get us to buy more and more expensively. But we don’t mind – in fact we lap it up.

The difference, this research tells us, is in how the predicting company is perceived by the shopper. If you come across as being competitive, trying to get that extra sale, then your shoppers will try to fool you and will avoid as much of this upselling as possible. However, if you are perceived as co-operative, helpful, supportive and as a partner, then we don’t mind being predicted – in fact we quite like it.

It all comes back down to one thing ultimately – personal service. Online, just as in the physical world. we prefer to be treated in a friendly, co-operative manner. Treat us like numbers and we’re off. So, it begs the question – how do you treat your online customers? Do you personalise your service, give them a sense of co-operation and helpfulness? Or do you simply treat them as a source of cash – and hopefully some extra money as a result of predicting those extras they will buy?

If your website falls into this last category you can expect that even though you cannot hear them most of your visitors are shouting at you rather like Adam Boulton.

Like this article?

Share on Twitter
Share on Linkdin
Share on Facebook
Share via email

Other posts that might be of interest