Is Targeted Advertising Creating an Uncanny Valley?

Marketers are increasingly engaged in a technology arms race, finding ways to derive ever more intimate details of consumers so that advertising can be targeted with precision accuracy.

Because of the positive ROI implications of targeted advertising, we are now seeing brands finding ways to turn what were previously offline activities into data. So, for example, at one stage Tesco was piloting Amscreen’s facial recognition software at their petrol pumps enabling advertising to target advertising based on personal characteristics such as age and gender.

But just how comfortable are consumers with these sorts of activities? Amid the excited talk of algorithms and programmatic advertising, it is easy to forget that at the end of the chain sits a sometimes emotional, often irrational and typically unpredictable human. And they are not always convinced about the way in which technology is being used to target them. Survey after survey report that consumers find it creepy the way that brands currently use information held on them. And perhaps a tangible manifestation of this is the huge growth in online ad blocking, sending a very strong message about their unwillingness to receive personalised marketing.

It may be possible to claim that these are merely the symptoms of a mindset that is adjusting to new forms of advertising. After all, it can be hard to find people that warm to TV advertising while there is a long history of research demonstrating the consumers both accept it and are influenced by it. However, is there something different about advertising that targets consumers based on their personal details? It is a question worth asking as there is a growing debate whether brands risk peering into an ‘uncanny valley’ where consumers come to reject highly personalised marketing approaches.

The Uncanny Valley

The term ‘uncanny valley’ was first used in 1970 by Japanese roboticist Masahiro Mori, who noted that although we tend to warm to robots that have some human features we tend to be disconcerted by them if they start becoming too realistic. And while there is little empirical evidence to support this claim it has nevertheless gained traction and steadily been gaining momentum ever since.

The uncanny valley effect has since been blamed for the failure of a number of films that used CGI where the characters have been very human-like while the audience is aware that they are in fact animations. The Polar Express is often cited as an example where the effect left it with lackluster box office sales whereas films such as Brave or The Incredibles used characters that were clearly not human and fared much better. Circumstantial evidence to be sure – but interesting speculation nevertheless.

Information Imbalance

So do brands face their own uncanny valley of personalised advertising? In principle, we may well expect to see the same issues that Masahiro Mori found, given that this advertising is via computers generating increasingly personal interactions based on ever more intimate knowledge about their targets. From a brand relationship perspective, it perhaps reflects the way in which personalisation creates a basic imbalance between the players. So while the brand gets increasingly personal with the consumer based on all manner of information it holds, the consumer is unable to generate the same intimacy with the brand, which can lead to suspicion and a lack of trust; just as we are likely to be suspicious of people we barely know who seem to know all about us.

This may then explain why we tend to find such negative reactions to personalised marketing. A possible example of the uncanny valley in action comes from Qantas. Flight attendants now have very detailed data on the airline’s highly-valued frequent flyers displayed on their onboard tablets. Despite this being something actually sought by Qantas staff they found it difficult to incorporate this information into their interactions in a way that felt natural. So instead of making their most valued customers feel looked after, this data-driven approach too often ‘creeped them out’. And, in another example, Urban Outfitters attempted to personalise their site so that the website’s product displays to match a user’s gender. Instead of appreciating this move, it came too close for many of their consumers’ comfort.

Research suggests there are optimal points of personalisation beyond which consumer response moves from positive to negative. And that point may well vary depending on a variety of factors. Some categories may be more associated with these approaches – what is appropriate for Google may not be right for an FMCG brand, for example. There are also very clear differences in receptiveness to targeted advertising across different population segments – partly but not wholly based on demographics such as age, gender and lifestyle.

Marketers have enthusiastically adopted the mantra that personalised marketing is the right strategy for their brand. And while much of the time this may well have much to recommend it there is currently no understanding of the optimal way it should be deployed. But it is increasingly looking as if marketers need to start asking where the uncanny valley begins for their brand. Because your brand might just be doing the opposite of what you intend and actually turning consumers off through your activities.

By Colin Strong