Emotional Intelligence: Balancing data and creativity

Where does creativity exist when everything becomes automated and targeted? Are we in danger of stifling outcomes with predictions based on our prior interactions?

In Spike Jonze’s Her (2013), the protagonist Theodore Twombly, an introverted writer and gamer on the verge of divorce, purchases the first artificial intelligent operating system, OS1, to literally run his life. During setup, the operating system — named Samantha — takes on a personality that’s then grown through their interactions and experiences and the ‘OS’s’ hunger for knowledge. As Samantha becomes increasingly attentive, she and Twombly begin to fall for each other. Despite the fact that Samantha does not have a body, they come up with ways of evolving their relationship into one that feels fulfilling in all facets of the human sense.

Charlie Brooker’s TV series Black Mirror pushes similar technological hypotheses into darker terrain. In the episode “Be Right Back,” the central character Martha is devastated when her social media addicted partner Ash is killed in a road accident. A friend prescribes her a service that allows people to communicate with deceased loved ones by using their online contributions and interactions to help with grieving. Brooker’s idea is that software can build a “you” based on all your emails, texts, Facebook posts and Tweets. What’s of interest to this story is the singularly optimistic nature of online posts — we only share the good stuff. As a consequence, the programmatic version of the deceased Ash is not equipped to deal with Martha’s grief.

Both stories present an exploration of human need and interaction, but also an exploration of our relationship with the technological interface. Just look around you and see the faces of family, friends and colleagues illuminated by the screens of all manner of devices. A new breed of co-­dependence, whether it’s an insatiable appetite for knowledge, a desire for game world immersion or the vicarious pleasure of observing friends through their curated posts.

The way we engage with devices and computers has evolved. Touch interfaces and voice recognition are commonplace. Usability is increasingly intuitive. Our inputs are changing, from identity with fingerprint scanning and face recognition, to nuances in speech intonation. This increasingly detailed fragments of familiarity (let’s call it ‘user capture’) will ultimately produce data that’s never been captured at scale before.

Taken to extremes, this approaches the 1950’s notion of the “singularity” and “intelligence explosion,” espoused by mathematician and author of The Computer and the Brain, John von Neumann. By this, he was referring to the expansion of a device’s cognitive abilities to the point at which they surpasses that of any human.

It sounds like the stuff of science fiction. Yet the retailing giant Target uses customer tracking technology to identify key life moments based on complementary product purchases. Their statistical genius has created a formula to identify when a woman is likely to be pregnant from 25 products purchased together, for example. This then triggers the delivery of targeted coupons. And the imaginings of Jonze and Brooker is only fractionally ahead of Amazon’s suggested products and recently patented “anticipatory shipping”, a feature using their truckloads of customer data to predict what product you’ll purchase next. For Amazon, this will cut down delivery times and put them ahead of their real world competitors. For us, it serves as a reality check as to how well computers might know us.

Similarly, Google is busily amassing not only Big Data but “truly massive data… from the minutiae of our lives.” Google Now takes your world of ‘Google’ and presents back to you on a neat little card the most likely thing you’ll do next: a card for the weather, a card for traffic on your route to work, a card for your next meeting, the footy scores and a reminder of a friend’s birthday. There’s no doubt that brilliant service design keeps you coming back for more, in the way that a real world barista who knows your name and order will see you coming back each day.

However, such acuity can be kind of Aspergic — like Martha’s new “companion”, it does not always allow for the emotional nuances of human situation. In the case of Target, for example, the system in one case identified a teen pregnancy before the woman’s parents did (the teen’s father found coupons in the family mailbox for cots and baby clothes) — insert “whoops” emoticon here! This caused some confused and heated debate between the woman’s father and the store, and even more discussion at home afterwards.

So herein lies the challenge: it’s true that pattern based predictions can be helpful. Data affords us an incredible insight into users and their behaviours. But we’re in danger of limiting the future by the actions of the past. More importantly, we’re in danger of cramping our creative style.

Creativity, by nature of the word, involves a level of originality and surprise. It not only plays on what has come before (ideas, texts, genres, interpretations) but seeks new ground. It is not only the track beaten down behind but the rocks, crags and waterfalls ahead. Creativity, the good stuff, takes us further than we’ve known.

Anita Elberse is a Harvard Professor whose book Blockbusters: Hit-­making, Risk­taking, and the Big Business of Entertainment investigates the strategy of blockbusters in entertainment. After analysing a decade of data she clearly identifies a formula by which movie execs are minimising risk and maximising return. They go with what’s known in terms of both the story and the cast. This means sequels and book adaptations (read: demonstrated audiences) in plentiful amounts. 2015 is rather telling with The Avengers Sequel, Batman vs Superman, Independence Day 2, Jurassic Park 4, Star Wars VII and Terminator 5. But is there a danger that the formula will simply wear out?

In advertising, the answer is yes — and with digital data, as with box office, these failures are self­evident. In a cluttered and savvy media world, lacklustre messages fail to convert, to attract, to engage, and this activity (or not) can be easily tracked.

In the case of programmatic paid media, advertisers have the ability to chase users around the web until they succumb to a funnel of conversion and then purchase. But however optimised that experience is — from the specifics of button colours to exhaustive multi-variant testing — the ad’s effectiveness will wear out with exposure. They will simply stop being noticed.

We know that putting the user first ensures the experience works better. Now with native advertising (ad content overtly fused with the user’s experience), advertisers have to think about their messages as branded content of specific use or entertainment to the user. In the case of editorial native advertising — you might think BuzzFeed, but know that The New York Times et al have offerings too — the publisher plays creative partner with the brand and they work together to create content that is mutually beneficial for both the reader and the brand.

Mini, for example, partnered with BuzzFeed to drive its ‘Not Normal’ brand positioning; a partnership made in heaven due to BuzzFeed’s audience hunger for the bizarre and abnormal. A dozen articles celebrated “Not Normal” via fashion oddities to environmental phenomena. This content alliance allowed Mini to expand its brand messaging positioning and bring their range of unique cars into focus.

In the case of social native advertising — think sponsored Facebook posts and promoted Tweets — the page or profile owner needs to be clear on what value their brand brings to their followers and not veer from that purpose. More telling in this instance is the reactions to the content — was it referred, commented upon, liked etc. Actions validate the content and give the advertiser a unique perspective on where their storytelling efforts should go. It also gives them permission to scale via sponsored media to replicate the endorsement that social uniquely gives.

Oreo’s ‘Daily Twist’ campaign is a good example of the US cookie brand knowing what they stand for and where to fit into social. One hundred ads in 100 days turned trending news stories into visual treats. The already topical stories had an audience from the outset; Oreo simply repurposed them with their twist, then sponsored the content so their creative became news in its own right. Of the hundred posts, the most famous, ‘Dunk in the Dark’, seized the moment when the Super Bowl broadcast blackedout, the created image was shared on Twitter and Facebook more than 20,000 times and was estimated to reach five times the number of people who tuned in to watch the game. A planned approach meant Oreo concentrated its efforts on relevancy, reach and timeliness with an outcome everyone talked about.

It’s evident that programmatic media can be relevant, but it lacks soul and context, whereupon native formats are reigniting creativity from both the brands and the publishers to engage with audiences in more interesting ways. Native also affords referral and is more embedded into the social psyche of the web. I predict that programmatic media will be the death of what can be only described as traditional online advertising, or The Banner Ad. Yes, a powerful partner, but misused it turns into a creepy stalker.

On the one hand, it’s like painting by numbers; on the other, Jackson Pollock. Is there a happy middle?

Data isn’t necessarily devoid of emotion. It can be creative in solving problems and, in some cases, even revealing great beauty.

In the United Kingdom, the Greater London Authority (GLA) has created The London Datastore, launched in 2010 by the Mayor of London Boris Johnston. It’s a free and open platform that gives citizens access to public sector data. It’s a smart move in citizen empowerment as the GLA appreciates that raw data often doesn’t tell you anything — it’s not until it has been presented in a meaningful way that unforeseen outcomes can arise. They are encouraging technical talent to transform rows of text and numbers into apps, websites or mobile products which people can actually find useful.

The GLA initiative has already given birth to some interesting projects: from the fun and educational London Jigsaw App, where users get to show how well they know London by rebuilding it one piece at a time;; to helping Londoners commute via the City Mapper, which not only helps them navigate public transport via access to ‘Transport for London’ data, but prompts them to consider walking or cycling by showing how many calories they’d burn. Being open with their data has allowed London to rally its technically savvy citizens to create tools and services that improve the lives of many. Open data in the context of cities has amazing opportunity to create understanding around societal problems and potential solutions. By cross referencing data on crime, geography, education, demographics, ethnicity etc. governments and citizens can start to understand the patterns that lead to problems.

Where data begins to approximate art is in the hands of artists like Aaron Koblin and Jonathan Harris, whose timeless examples give a unique perspective of the world via data. In Koblin’s early work Flight Patterns , North America is reinterpreted with captured flight 5 data. With no reference the continent takes on a new form via the air traffic of its inhabitants. Harris’s We Feel Fine observes humans in another way, by exploring emotion 6 of the published kind. The software harvests sentences containing the phrase “I feel” or “I am feeling” from newly posted blog entries (note: this was before Twitter and Facebook). It then displays them in a compelling interactive experience where you’ll lose hours exploring it. In another observation “I Want You To Want Me,” an installation commissioned by MOMA in New York, Harris explores the search for love and self in the world of online dating. The installation chronicles the world’s long-­term relationship with romance, across all ages, genders, and sexualities, using real data collected from dating sites every few hours. It feels to me that these sensitive and thought-provoking pieces shone a light on the possibilities of connecting with audiences in radically different ways. Yet are now silenced by targeting technologies that chase you around the web. Can a balance be struck?

Vocativ, the global social news network, uses data-­mining software that was targeted as a tool for law enforcement and government agencies. The software, OpenMind, accesses the “deep web” which standard search engines overlook. Content like spreadsheets and Word documents and dynamic pages. Vocativ claim they can use the deep web, in combination with monitoring social media, to find news stories other agencies can’t. What makes Vocativ’s approach to the newsroom unique is the pairing of journalists and data analysts to produce stories. The technology helps discovery and investigation then journalists and analysts curate and create stories. This is a smart balance that respects the craft of persuasive writing, yet uses the brawn of data mining to go beyond the surface of the web.

In my own work we deployed social web spiders to index and identify specific audiences and their behaviour with content. The aim was to identify what engages them most so we could build better content experiences. The audience was C-­Level decision makers within the Information Technology space. This audience is information hungry and frequent social channels to glean the latest opinion and news on their area. By using directory listings we were able to identify the top companies, via LinkedIn the key decision makers, then matched their profiles to Twitter and other social profiles. From there we were able to build a unique topic dictionary that was tailored to this audience’s interests then use it to highlight how they behaved with industry content; news, white papers and video case studies. Knowing what worked the best allowed us to tailor our own content efforts and most importantly measure it once it was released to this select audience. We were able to see who and what influenced them and create opportunities to expand stories and ultimately connect to drive sales.

It feels that we are at the intersection of change, whereupon push advertising lacks a role in our online experiences. We naturally look for referred media first, something that’s validated by real people, ideally our friends.

We’re never going to refer a well targeted banner ad that happens to know I’m in the market for a new coat. We won’t even celebrate the search engine optimised website we found as its likely geared for search bots not human eyes. We will, however, refer friends to a remarkable film or story. We’ll encourage followers to download an App we can’t live without or visit a site that’s interaction design is compelling. We’ll share information that’s helped us solve a problem or tackle a task we knew nothing about. We will even applaud a service that took the time to look after our needs and go the extra mile. Experience is everything and if you’re having a good one, you’ll likely shout about it.

So the data we create knows a lot about us. Sometimes more than we do. But we must all find a balance where what it tells us is not the rules but the opportunity to create, to connect. Creativity remains as important as ever to provoke reaction and interaction. We need to find that sweet spot where a hyperlinked virtual and (with connected homes and wearable devices) now physical world can actually help our stories and information connect. It’s true that companies and brands will be enticing us to share more of our behavioural data for benefits.

Ironically, in Her, it was Twombly who was stuck in a formula, and it took a machine — albeit one with the seductive voice of Scarlett Johansson — to teach him how to “live”. By stepping away from his devices, rejoining the humans and learning to risk again.

The takeout of this is not necessarily just a philosophical one, however. The “formula” in which advertising often finds itself stuck is one of structure and process. If programmatic advertising is the lot of media companies and emotional storytelling that of creative, what is needed is in fact better communication practices between the two. As media companies begin to forge creative capabilities and creative companies become more media savvy, we edge towards the brave new world of an agency capable of a holistic approach with laser accuracy and intuition. The irony is, this is where we came from albeit when media was simpler before the industry fragmented and we became, by some degree of necessity, specialized. But now we need to return, head and heart joined and ready to take a risk.

This essay was first published in Hacker Maker Teacher Thief: Advertising’s next generation by Creative Social. I run the Sydney chapter of Creative Social.