Feature

Is there an algorithm for creativity?

Algorithms that mine our data are said to know more about us than we know ourselves. Nicola Kemp asks what the implications of this might be for brands

Is there an algorithm for creativity?

Algorithms are a force of nature. In an age when everything, from architecture to the jobs we are offered, the films we watch and the books we read, is governed by algorithms, their all-encompassing power is increasingly difficult to ignore. For marketers, the intersection of big data and technology is providing unparalleled insight into consumers’ behaviour – a shift that some believe could lead to a revolution of the entire creative process.

It is widely accepted that creativity cannot be emulated by machines, but are we naive to believe this? The advertising and marketing industries are at pains to protect the processes by which creative thinking is born, but can the creative process now be automated?

It is widely accepted that creativity cannot be emulated by machines, but are we naive to believe this?

People who define themselves as "creative" find it intrinsically difficult to take an objective view of the power and possibilities of algorithms. However, the fact is that an algorithm could legitimately write this article and, arguably, do a better job. US company Narrative Science has built a set of algorithms that build well-styled and grammatically correct sports reports based on the data related to each game; they are used by a range of media companies, including broadcaster Big Ten Network. By this logic, algorithms can be more productive than any journalist.

Christopher Steiner, author of Automate This: How Algorithms Came to Rule Our World, says the story of the next 20 years will be the tale of big data and algorithms. "We are at a giant fork in the arc of humanity [and the question is] just how much will we allow algorithms to take over?" For marketers, it is a question that cannot be ignored.

Big data is dead

In many ways, the rise of the algorithm is a reaction to the empty rhetoric of big data. "Big data is dead; gathering information for information’s sake is not enough. We need to make data useful, and that’s where algorithms come into their own," says Jerome Courtial, head of strategy at digital agency We Are Social.

In the marketing industry, the rise of the algorithm is all about a new breed of intuitive technology that understands consumers’ buying habits and makes relevant recommendations. While computer automation should not be seen as a one-size-fits-all solution to the data mountain, many of the world’s most innovative companies, from Facebook to Netflix, have used algorithms to better collate behavioural data as a means to improve the way they target their consumers as individuals.

Algorithmic autonomy

"Mihkel Jäätma, managing director of Realeyes, a company that measures facial expressions and quantifies consumers’ emotional reactions via devices such as webcams, says: "We can measure emotion on a large scale and match this with performance data to predict business outcomes. As the technology becomes more adaptive and intuitive, it will become a bit likeMinority Report, as we will use more live emotional feedback."

Now that marketers can effectively plug into "a Google for emotions", not to mention the ever-widening range of software available, they are embracing the power of the algorithm. James Kirkham, co-founder of digital strategy agency Holler London, says algorithms are playing a greater role in the creative process, with a growing number of agencies using algorithmic analysis to underpin the creative process. 

He explains: "Where once the creative sat aside from those who were listening and analysing, only being briefed at a later date, now the creative sits alongside those listening to communities, mining data and forming thoughts based on analysis, interaction and algorithms. Speed of response is now so vital that the algorithm at the heart of the cycle can trigger a creative cue that will be tested and put up online in a matter of minutes." 

The human touch

It is a well-established fact that people are poor predictors of their future behaviour. The notion that an algorithm could be a better indicator of their future purchase decisions than they are is difficult for some to fathom. However, the human tendency to "post-rationalise" behaviour is a trait that algorithms do not suffer from.

It’s a really dangerous precedent to say that marketing is all about the data and algorithms. All the best companies are constantly in contact with their consumers.

This is not to say that algorithms are bulletproof. Courtial believes that while there is an important role for algorithms in marketing, they come into play only when consumers already have a relationship with a brand. "We need to recognise the limitations of algorithms; for example, when you make a one-off purchase like an electric cable, Amazon keeps emailing and updating you with suggestions of more electric cables to buy – it’s off-putting," he says. 

According to Courtial, herein lies the fundamental flaw of algorithms in marketing: namely that they are great at studying consumer behaviour in the past, but not so reliable when it comes to predicting future behaviour. 

Controversially, more and more software providers are privately marketing their products on the grounds that brands can reduce headcount if they place a greater reliance on automation and algorithms – a strategy that many in the industry believe won’t work.

Michael Veitch, managing partner at creative agency Rehabstudio, says there is nothing more important than human input, because only people can bring randomness, spontaneity and personal insight to the table. "Most algorithms work by suggesting things you might like based on what you already like; they don’t account for something you might not currently like, and will have to force yourself to, but will one day love," he adds.

There is no doubt that algorithms provide a fantastic tool for marketers, but some marketing experts believe they have yet to truly unlock the psychology of consumer behaviour.  Andy Bird, co-founder and executive director of Brand Learning, says the key to successfully embracing data and algorithms is balance. "It’s a really dangerous precedent to say that marketing is all about the data and algorithms. All the best companies are constantly in contact with their consumers," he explains. 

The illusion of choice

Of course, there are also myriad ethical issues surrounding the seemingly inexorable rise of the power of the algorithm. The biggest concern is that algorithms, while offering consumers the illusion of choice, are potentially robbing them of chance encounters, the element of surprise and serendipity that makes life interesting.

Anna Foster, data director at TMW, says there are legitimate concerns to do with removing this element of chance. "This is why most people agree that a world entirely governed by algorithms is leading people down a narrow path of homogeneity; it is almost anti-human," she states.

Questions remain, then, as to whether the plethora of algorithms at work in consumers’ everyday lives are effectively combining to create a narrowing world view. "When you type something into Google, you think you are getting the very best answer, when in fact you are getting the answer that best fits with your search history," says Courtial. 

He adds that the algorithm is effectively denying people the act of genuine discovery, which, he argues, is a fundamental human need.

Algorithms killed the marketing star

So will brands effectively outsource their core marketing functions to algorithms, and can an algorithm replace flesh-and-blood marketers? There is evidence that so-called "digital first" businesses have already outsourced traditional portfolio-management tasks to algorithms. 

As online and offline worlds merge, better deployment of data and algorithms to drive sales on all platforms will certainly be a focus.

Chris Gordon, business development director at digital marketing agency Ingenious Rapport, says that while even a brand such as Netflix needs to advertise to encourage consumers to subscribe in the first place, it can then use an algorithm that "does the job of an old-school brand manager in directing [customers] to new series and managing the portfolio".

The tipping point will occur when traditional FMCG companies such as P&G properly embrace this technology. "P&G could use algorithms to blur the line between physical and virtual spaces and direct [consumers] to its product rather than the competition," adds Gordon. "At a time when brands are trying to trim staff numbers, using data and algorithms to drive purchases will be key." 

Elsewhere, brands such as Tesco and Ocado already have huge volumes of consumer data at their fingertips, and the retail industry as a whole is making greater use of online shopping data to inform the decisions they make on the high street. As the online and offline worlds continue to merge, better deployment of data and algorithms to drive sales on all platforms will certainly be a focus.

There is no doubt that marketers are on the cusp of a transformation in business practices. "Anyone who says data and algorithms are unimportant and that it’s all about creative hunches will get left behind," declares Bird. "The truly great marketers will join up data and creativity and realise that they aren’t mutually exclusive pursuits."

Can an algorithm have a soul? 

There remains a cultural tension in the marketing industry between art and science. While data scientists may now find themselves revered by the industry, many believe the algorithm alone is not enough to truly drive creativity. "The human creative part will always be necessary, as you can’t seduce a consumer on data alone. We’re still emotionally driven creatures who need to be seduced and our souls enriched," says Kirkham.

Some in the industry believe that the very efficiency of algorithms is their shortcoming – that the common sense inherent in programming could never deliver Unilever’s Marc Mathieu’s "more magic, less logic" approach to marketing.

Martin Greenbank, head of intelligence at Arena Media, is sceptical. "Isn’t what makes us different from animals our cognitive ability to foresee and project abstract outcomes? So, unless we develop some seriously advanced artificial intelligence, there will always be a need for creative types who can dream up moonwalking ponies," he contends.

In many ways, the shortcomings of the data industry in selling algorithms on the basis that they can reduce the headcount in a marketing team highlights the inherent problems in an approach that places numbers first and people last.  "Organisations are made up of people, and consumers are increasingly demanding that they become more transparent and authentic. If you throw out this human element because you think you can do it with an algorithm, the risk is that the human touch, and the strategic thinking that goes with it, will be missing," says Bird.

Algorithms are indeed a force of nature, but like any part of society’s complex ecosystem, they cannot exist in isolation. The very best marketers understand that creativity and science are not mutually exclusive.