When personalisation goes bad
A view from David Paice

When personalisation goes bad

Personalisation isn't right for all brands and should be used with care, warns David Paice.

There are only two reasons that humans do anything - to move toward pleasure, or away from pain. These are the basic primordial instincts that guide our actions. Personalisation is probably aimed at supporting the former, but too much might push people to the latter.

Getting it right is not easy - and even when you do, personalisation might stray into territories that are deeply ethical or religious. Gene-splicing, designer babies, personalised drugs - these innovations polarise society and push the view that the debate on personalisation is being led as much by the technology itself as by consumer preference.

The trouble with technology driving the innovation is that it becomes a solution looking for a problem.

Research company Gartner's 'Hype Cycles' - intended to help interpret technology hype by showing the maturity and adoption of tech and applications, and how they can be relevant to businesses - currently plot personalisation within the "slope of enlightenment", suggesting that marketers slowly understand what is and isn't useful personalisation in their markets.

In the visitor-attraction industry, the opportunity to personalise communication to audiences takes many forms, some of which might even be considered hygiene factors: addressing emails to people by name rather than "Dear Guest" is basic manners. But do you use the first name or title and surname - what is the cultural norm in their region?

Successful personalisation has its roots in data science and analysis

Similarly, sending guests' photos from their theme-park rides to our retail stores so they can print them on anything from mugs to puzzles is a very popular opportunity to personalise their memories in a physical form. Both of these cases are considered appropriate because they are completely relevant to the context in which they are applied - they are an expectation, not a surprise.

This is important. It informs us about when personalisation can be used to bring pleasure instead of a nasty surprise. Consider the famous example of the US father who was furious to find his high-school-age daughter targeted with coupons for baby clothes and cribs. The retailer used her purchasing data to correctly infer that she was pregnant before her father knew.

For those wondering whether to personalise or not, the starting point is not what can be done with the tech available. Instead, it is what can be inferred from the data on your customers. Successful personalisation has its roots in data science and analysis. Tools that support success in this context are aimed at driving additional insight into your audience - Mosaic profiling, drive-time analysis, average spend and dwell times. These will give you far richer insight into your customers and offer far more successful cases of personalisation.

Even then, the technology has some catching up to do. I am amazed by how many online retailers re-target advertising to me displaying the exact products I have just purchased from their sites.