The paradox of marketing personalisation (a long post)

A lot’s already been said about personalisation in marketing.  To be honest, despite all the talk, I’ve so far seen very little of any marketing that’s truly personal, relevant, useful and interests me on a personal level.  But it’s clear as an industry that’s where we’re headed so we need to be prepared for it.

But what strikes me about the current discourse, is that it doesn’t feel properly thought through.  It’s as if, because it’s possible, we should just go ahead and do it and think about the objective later.  And what if, the outcomes don’t meet expectations or even worse, personalisation is having the opposite effect and leaving people feeling isolated, driving even more distance between people and brands?  I think that’s worth exploring.

Before the digital era, mass marketing was quite unsympathetic to the individual.  Audiences were lumped together and traded as commodities through blunt broadcast media.  Everyone was largely the same, distinguished mainly by age, gender and income. But digital media should now allow us to identify and empathise with the real people, the real individuals that actually buy our products.  The algorithmic systems, artificial intelligence and programmatic ad technologies that can fire laser-targeted messaging missiles at individual customers are pretty powerful weapons.

But ‘with great power’ as Uncle Ben once famously said to his future superhero nephew, ‘comes great responsibility’.

The ‘great responsibility’ for marketing is to stay true to Ted Levitt’s maxim that marketing ‘is an integrated effort to discover, create, arouse and satisfy customer needs’.

The real danger we face today is in wielding the power of personalisation to satisfy the needs of marketing, and not the customer’s.  And the damage has already begun to show.  The personalisation paradox is upon us.

Japanese roboticist Masahiro Mori is often credited with coining the term ‘Uncanny Valley’ in the 1970s.  It describes the revulsion people feel when they experience robots behaving almost, but not exactly, like humans.

In marketing, this happens when digital experiences present a likeness to our needs and wants that’s somewhere between being too close and not quite close enough to what we already know about ourselves. When we know we’re being targeted personally but the content or the data that feeds it doesn’t match our understanding of ourselves, the uncanny emerges.

We’ve reached the point people are now quite literally taking out restraining orders against marketers in the form of ad blockers. It’s really no wonder 99% of consumers don’t trust advertisers with their data[1]. These warning signals are all completely of our own making.

Is there a possibility that personalised marketing’s clinical, reflective and intimate delivery leaving people feeling cold and alienated is creating a relatively less human and therefore less personal form of connection?

This is the personalisation paradox: personalisation that fails to arouse and satisfy the human condition, which as a result becomes even more impersonal.

As Levitt said, marketing needs to start with the human.  Humans are irrational, social creatures carried by emotions and feelings that need to be understood, aroused and satisfied. Too many examples in this new era of personalisation are self-serving to marketing, not the needs of the human.

OK, with more data and more time, personalised marketing and people will eventually learn to get along.  But this is a simplistic, convenient notion.  Instead, personalised marketing must overcome neglect for the human condition by using marketing in a much more, well, human fashion that matches our understanding of ourselves.  How can personalised marketing avoid the the pitfalls of becoming impersonal and dehumanising and actually gratify our most human motives?

Connecting the individual to social contexts

Personalised marketing’s foundations are built upon the belief that we are independent, conscious decision-making creatures.  This belief that the individual has primary agency of behaviour is what Mark Earls calls ‘the illusion of I’[2].  The truth is that we are social animals influenced by the tribes we move in.

Marketing, whether personal or otherwise should be effective in the social contexts we live in and interact with. We have evolved to develop complex social behaviours and we are endowed with a social brain that understands, empathises and mimics other humans.  Feeling connected to and being influenced by others is an inherently human condition.  We are an ultra-social species for whom other people are the most important part of our environment.  Earls also says that consumption is also a social act.  What we buy and consume is influenced mostly by what other people do and think.  This idea has served marketing well.  Brand logos tend to face outwards for a reason and as proven by Binet & Field, Fame, or people talking to each about your brand, is a key principal of marketing effectiveness [3] .

So can personalised marketing reconcile with the hard-wired social identities of the ‘we-species’?

Yes, by personalising connections to others.

Data from activity on social platforms can provide insights into our interests, social circle, life events, family members, where we eat or hang out, (when and who with) and much more about our social lives.  There are obvious opportunities to target within these contexts.  For example targeting new mothers identified through their social footprint with baby products is an obvious easy, zero-wastage win for marketers.  But this misses a much bigger opportunity to connect the individual to other people or to a more human, social context.

Facebook data suggests that new mothers are most active between the hours of 04.00 and 07.00 as they nurse their restless babies at unsociable hours through the night.  But rather than simply serve product messaging at them at this time (which in itself is implicitly isolating), why not connect these mothers to others who are online at the same time to share their concerns, help and advice?  Brands have the opportunity to use personal data to unite people and create a sense of belonging to benefit the customer and the business.

Nike+ tracks users’ personal running data, creating personal goals to beat and dashboards to track progress.  But crucially, the individual’s performance is benchmarked against the entire Nike+ community and friends identified through the user’s social graph.

Personalised marketing can also have much greater effect if counter-intuitively; we look beyond the individual and apply its benefits in the wider social context to create a collective sense of value and ownership to customers or citizens.

Waze is a good example of scaling individual driving data to offer intelligent real time driving instructions to beat the traffic for everyone, not just the individual driver. The significance of apps such as Waze to marketing goes beyond the opportunistic temptation to sell more stuff to people based upon location.  Waze owns huge amounts of personal data that could improve the way cities and transport infrastructures are built.  Nest is another example of a business that can scale individual behavioural user data to influence local community and government energy policy.

A customer’s personal data could be used by brands willing to step into spaces the state is increasingly struggling to fill by offering solutions to wider social issues.

Leading with emotion and the gratifying the storytelling self

Behavioural targeting and the ability to retarget customers along their purchase journeys has served the marketing industry well.

That’s not to deny all the consumer benefits to behavioural retargeting too (being served relevant, useful messages often with the ability to refine or opt-out), but outside of marketing circles, being chased across the web to ‘buy now’ is often considered at best intrusive and a bit of a nuisance.

More importantly to marketers wishing to overcome the personalisation paradox, persistent cold and clinical ‘buy now’ messaging is not only a wasted opportunity to respond to human emotion beyond interest and action signals, it also fails to satisfy the human desire for storytelling.

Humans make decisions based on feelings, emotions and context.  It’s Kahneman’s System 1 brain that predominantly decides what we spend our money on[4].  New technologies now enable us to read individual human emotion and react to it in real time to offer increased relevance to the customer, specifically deep learning systems and biometric technology.

Deep learning systems are designed to understand the sorts of things computers have traditionally struggled with but humans find easy.  Emotions, facial expressions, language and speech for example are all things that humans process intuitively but can’t really define how (could you explain how you recognise yourself in the mirror?).  Deep learning systems effectively teach themselves the rules their programmers can’t.

In 2015 Google bought DeepMind for $400m to better understand human emotions, speech, language and thoughts and Google’s Deep Dream project is aiming to teach computers to see, understand and appreciate the world the way humans do through image recognition.  Facebook’s DeepFace technology can recognise specific human faces in images around 97% of the time, around the same level as humans.

And natural spoken language is already a prevalent feature of the technology powering Amazon Alexa, Google search, Apple Siri and Baidu’s Duer to assist us in everyday life.

Developments in voice and image recognition is significant for marketing personalisation because we will start to understand more specific, nuanced, conversational, contextual and emotionally driven needs and wants, instead of following people around the web based on historical web browsing behaviour.

To truly understand what customers are feeling or what they want instantly and consensually, perhaps the future is in biometric data.  Biometric data through wearable technology or through personal devices are already being successfully used in other industries from healthcare to gaming.

Mood, location, presence of others, physiological responses, movement, temperature, facial expression – the list of data points and the combination of these data points is endless, opening up real-time activated marketing opportunities that could be based on future behaviour rather than historic.

Perhaps the greatest opportunity to overcome the personalisation paradox in the area of programmatic ad-serving is in rewarding the human thirst for storytelling.  Stories are the way we humans make sense of the world, they provide us with a sense of belonging and help establish our identities. Just like our language instinct, we possess a natural born ‘story drive’ which every culture tends to in order to explain how the world works and to engage emotions. It’s fundamentally in our nature to need and tell stories.

Universally, stories commonly share a theme of a protagonist overcoming a problem or challenge.  What if personalised marketing could make the customer the protagonist in her own brand story?

Programmatic delivery offers marketers the opportunity to tell a personally relevant brand narrative across multiple platforms, devices and touch-points, yet examples of creative programmatically served storytelling are scarce.

Can brands not only play out personalised narrative across time and devices, they could also create stories with multiple protagonists, stories to be pieced together by friends, thereby once again using personalised marketing to better connect to other people in our social environments.

Surprise! Serendipity and discovery feed the human spirit

The human brain is inherently lazy.

We like to follow the path of least resistance to preserve precious energy.  Cognitive decision-making is an energy-sapping endeavour so it’s a good thing that technology is continuously making choices for us.

The way we search, shop and consume is increasingly algorithmically pre-determined by technology that works out what we want often before we do ourselves. Amazon recommendations, Spotify playlists, friends Facebook posts and even your potential Tinder matches have been personally and intelligently calculated based on your previous interactions and behaviours.

Programmatic ad technology, the power of the cloud and ubiquitous access to it, are ensuring the world we experience is ever more relevant to our wants and needs.

But does this increasing control and predictability in our lives come at a price?

Whilst the brain is indeed lazy, it also thrives off surprises.  Surprise is an incredibly addictive drug that when confronted by a pleasant, serendipitous discovery, our brain’s reward pathways fire up.  Surprises also amplify our emotions, so if we’re feeling happy and surprised we feel delighted, creating positive neural associations with the stimulus, or the brand.

Surprise or a new discovery also has the ability to change behaviour.  Surprise introduces us to new information, new ideas and brands that force us reconcile against our existing beliefs and behaviours, creating cognitive dissonance.

Rather than continually reinforcing existing beliefs and predicted behaviours through the algorithms that power our digital experiences, perhaps we should be personalising the surprise to trigger a reappraisal amongst non-buyers.

Which follows onto the marketing case for providing customers with personalised serendipitous experiences.

Of course, identifying hand raisers and audiences most likely to convert makes complete sense in the short-term, but as we know brands grow by extending reach and attracting more people into them.  If our digital experiences are increasingly trapping us in to our personal filter bubbles, the ability for marketers to expand their customer base diminishes.

Algorithmically curated personal recommendations presuppose the machine knows what the customer wants.  But technology also gives control back to people to allow them to curate their own choices like never before.  Pinterest, for example is based entirely upon human curation. And Apple has put its faith in human editors for its new music streaming service and Beats 1 radio station to help people discover new music that computer programs can’t.

Automation of choice has its place.  But serendipity; surprise and discovery feed the human spirit.  Brands that are able to balance the short-term gains from satisfying our immediate and most relevant needs with the desire to challenge our pre-existing beliefs and behaviours will find its way into the lives of existing and new customers.

If this truly is the era of personalisation, then marketing needs to adapt to put the needs of the human before the needs of marketing.  We’re already witnessing the effect of super-charged personalisation that fails to connect on a personal, human level.

To overcome the personalisation paradox we need to recognise customers as humans, as social creatures; driven by emotions with a brain hardwired for stories and surprises.  Only then will marketing discover, create, arouse and satisfy customer needs, which after all, is what marketing’s supposed to do.


[2] Herd, Mark Earls

[3] Marketing in the era of accountability, Binet & Field

[4] Daniel Kahneman: Thinking, Fast and Slow.