Will an algorithm generate the next big Christmas ad?

Christmas is coming and algorithms are at work, pushing the boundaries of creativity to ensure the tills keep ringing. But can AI tools ever really craft an ad that resonates on a deep, human level? Asks Paul Armstrong

The Christmas ads have arrived, including a fake – and viral – John Lewis one, flooding our screens with the usual mix of warmth, nostalgia, and all manner of emotional porn for the masses. Brands like John Lewis, Coca-Cola, and even local retailers have been polishing their emotional currency for months, hoping to leave a lasting impression and pockets lighter. But this year, there’s a shift. Behind the curtains of sparkling lights and mistletoe, algorithms are at work, pushing the boundaries of creativity to ensure the tills keep ringing. But can AI tools ever really craft something that resonates on a deep, human level? Or are we entering a future of formulaic sentiment that feels, well, a little too calculated?

The latest Cannes Lions showcased a new trend; AI-assisted creativity took centre stage. Coca-Cola’s 2024 AI-generated ad, produced in collaboration with OpenAI’s generative tools, bagged several awards. Ads were sleek, entertaining, and undeniably effective. Yet, you can’t help but wonder: was it touching the heart or merely pressing buttons pre-defined by data models? Emotional resonance is slippery. We understand the mechanics – certain visual sequences, music swells, and personal narratives drive emotional reactions. Behavioural economics has been breaking down these neatly for years. Are we now getting to a stage where these are programmable, repeatable, and, in the case of AI, scalable? Quite possibly. 

Can data decode emotions?

Consider tools like Daivid, the emotional AI platform that decodes the balance between emotion and effectiveness to help brands up their game at scale. The data says they can get people to buy more stuff. Suddenly, we could find ourselves looking at campaigns that are more about emotional predictability than genuine creativity. Ads that are optimised to evoke specific emotions on cue. It’s effective, sure, but does it move us forward, or just closer to a world where authenticity is lost, and we are left with formulae masquerading as art? To say nothing of sustainability and ethics. 

Welcome to the paradox the creative field is grappling with when it comes to AI. The fiefdom of insights is fast losing ground to untiring (if slightly still erring) black boxes whose ability to find patterns, to deconstruct human behaviour, to analyse the past in order to predict the future can be unnerving. Behavioural economics, at its core, is about leveraging biases – anchoring, loss aversion, social proof – and these are not impossible things to code. Plugging these into an algorithm is simple. And we are seeing the fruits of this approach: AI-generated content that works by all objective measures, hitting metrics, driving clicks, increasing sales. 

The human aspect of creativity often thrives on ambiguity, nuance, and the unexpected. The best campaigns – like John Lewis’ iconic Christmas ads – often work because they surprise us, even as they tug on familiar heartstrings

Yet, the human aspect of creativity often thrives on ambiguity, nuance, and the unexpected. The best campaigns – like John Lewis’ iconic Christmas ads – often work because they surprise us, even as they tug on familiar heartstrings. Ads that reflect the unpredictable, messy reality of human experience. Can an algorithm, by definition built on precedent, ever achieve that genuine element of surprise? More importantly, should we expect it to?

Coke’s upcoming holiday ad is all done with AI, but does it have the staying power to be remembered as a cultural moment like the original, or will it be just another artefact of algorithmic efficiency and huge media ad buy? 

Christmas ads provide a timely case study of where AI-generated creativity stands. There’s no doubt they’re getting smarter, slicker, and more efficient. But whether they can truly create something that shifts culture, rather than merely reflecting it back at us, is still very much an open question. For now, the algorithms can write a solid script, produce a visually impeccable scene, even predict what will make us laugh or cry. But to create something that makes us feel something new – that’s still the realm of human creativity, as messy and unpredictable as it is. 

Paul Armstrong, founder of TBD Group, helps companies navigate emerging technologies, to innovate and avoid disruption.