Remember the film, Groundhog Day? Bill Murray has to live the same day over and over again. Except the day is never exactly the same. Every time Phil the weatherman wakes up to the strains of Sonny & Cher’s greatest hit, he is able to subtly change things; accumulating more and more information about the people around him so he can make their lives better as well as his own.
You can see where we’re going with this, right?
At its heart, personalisation is about delivering a different experience to someone based on information you have about them. Whether it’s as simple as including a customer’s first name in an email or as complex as a completely different landing page experience for each unique visitor.
What’s important though, is not how detailed or complex the personalisation is, but how effective it is. It’s never easy to get that right, but across some industries it can be easier. For example, our work for Atom Bank was restricted to a certain set of behaviours, features, and requirements required by banking customers. That meant we were able to bring the ‘telepathic’ bank brand to life through a set of subtle and useful prompts and nudges across a reasonably finite user journey.
When you enter the world of entertainment, however, things become a lot murkier and less straightforward. What we’ve learned working with the likes of Channel 4, Warner Bros., J.K. Rowling and Vue Cinemas is that entertainment brands, products and customers require a very different approach. The patterns of consumer behaviour are wildly unpredictable, the industries involved move at light speeds, the diversity of the product is immense, and there is inevitably a long and crowded production process… That’s just the tip of the iceberg.
Of course the most basic level of entertainment buying is predictable. But a huge part of the entertainment buying experience is instinctive, impulsive and impetuous. It’s that serendipitous element of the entertainment experience that helps make it entertaining. If our tastes were predictable and ‘on rails’ then they wouldn’t be as much fun.
But that’s not much help when you’re trying to design a personalised user experience.
Unfortunately this tension has resulted in a lot of attempted ‘taming’ of that serendipitous element, with the wrongheaded aim of forcing it into a blunt algorithm. You will have seen this yourself, when a service has attempted to clumsily second guess your intentions and got it wrong. Followed by the inevitable overcorrection which involves intrusive questions and annoying added steps in the user journey designed to extract more ‘data’ from you.
Two names who’ve avoided this trap are Netflix and Spotify. These two have succeeded where others have floundered by embracing unpredictable user behaviours, unwieldy catalogues and exceptionally long longtails.
They do this by focusing on creating a level of trust and empathy with their users that attemptsto bypass the ‘If you liked this, then you’ll also like this!’ approach.
They avoid the temptation of clinical, blanket personalisation and trying to predict what your next move might be. Instead of trying to squeeze every last bit of personal preference from you, they look at smarter ways of fuelling their engines maximising their central brain with the knowledge of the crowds; and, crucially, they reflect those tastes back to you in a non-binary, expansive framework that avoids never being simply right or wrong.
At Th_nk we have begun to think of this model as ‘intelligent empathy’: the act of harnessing the power of human networks to create a semi-artificial intelligence that not only strives to get to know you, but works equally as hard to make sure you trust it.
That level of trust comes not just from repetitive use, but rather from increasingly more rewarding and meaningful interactions. Interactions which should be subtly curated, expansive, and immersive. And the fuel that powers those interactions has to be, to some extent, powered by other people almost as a by-product of their interactions.
Intelligent empathy is a virtuous circle: user behaviours fuel an AI that supports a recommendation engine that, in turn, is presented in such a way that the user is actively encouraged to interact with it, honing and perfecting it (instead of dismissing it outright). Those interactions become another source of fuel for the AI… and around it goes.
Which brings us right back to Bill Murray and that groundhog. Phil the weatherman can spend endless days acquiring knowledge about people and changing their lives in all sorts of ways. But it’s not until he masters empathy, really starts to understand what people need and having meaningful interactions with them, that he can be set free from the endless loop he’s stuck in.
This is an edited version that first appeared in The Drum Network Entertainment special
Rob Hinchcliffe, content & community director, Th_nk