The ‘Un’-Hidden Persuaders: How Data Changed Advertising For The Better

The Mars Curiosity Rover, at times as far as 250 million miles away from NASA, needs to be able to stop itself from falling down crevices, bumping into rocks or, who knows, even running over Marvin the Martian. But with a twenty-minute communication lag between the red planet and Earth, the lonely Rover must rely on highly complex autonomous hazard avoidance software (artificial intelligence, or AI) to keep the $2.5bn piece of equipment out of harm’s way.

Similar technology and data processing techniques have been quietly delivering groundbreaking advances in physics and biology for some time now, as well as in less apparent disciplines like gaming and advertising. In the same way that Curiosity can figure out how to navigate unknown landscapes without human intervention, virtual characters in games like Skyrim or The Sims can learn new behaviours, interact with each other and, essentially, develop independently.

So where does advertising come into this? Data has been used to some extent by advertisers and marketers to segment and target audiences for fifty years or more. So, there’s little new here. In fact, in 1957, American author Vance Packard went so far as to criticise the morality of advertisers in their use of data and motivational research in his book, The Hidden Persuaders.

While Packard may have been concerned with what he thought of as the manipulation of a vulnerable public, the use of data has allowed for a more personalised approach to advertising that offers people truly relevant products and services that they are genuinely interested in. Industry reports regularly indicate that consumers have a more favourable attitude towards targeted, relevant online adverts than many critics claim.

It’s fair to say that data is now the dominant force in digital advertising because, as advertisers have realised, it’s not just the number of the audience that counts. The ‘one size fits all’ advertising approach of Packard’s era has long since been proven defunct. Advertisers now must make complex choices in order to expose their messages to the right kind of people at the right time, in the right context. But, with such an abundance of data at their disposal, how can advertisers possibly achieve these results?

Those selling online ad space, or ‘inventory’ (such as contextual ads on search engine results pages, banner ads, blogs, rich media ads and social network advertising), sell their offerings via advertising exchanges, which price and trade in real time. Ad buyers can then effectively identify, bid and buy the ‘perfect’ space for their message and audience using live information, by the millisecond.

Via most ad exchanges, ad buyers have a selection of 12 audience attributes to choose from (ranging from typical demographic aspects like age and gender to online behaviour traits), with a total of 145,710 segments of data available on which they can base their bid decisions. Dizzyingly, the number of different possible combinations equates to over 504 septillion.

In other terms, if it took a buyer an optimistic ten seconds to review each option it would take 119,764,735,724,180,000,000 lifetimes to make that decision. Of course, this is completely unachievable so, instead, as with many industries before us, we are witnessing an industrialisation of digital advertising. By and large, humans are welcoming more efficient algorithms and artificial intelligence into their lives.

This is changing the way that marketers and advertisers work by freeing their time from repetitive data crunching, and allowing them to take a step back and learn from their campaigns; who is our most responsive audience, what does this mean for new product development and messaging, and what would happen if we tried something new?

In many ways, AI is allowing advertisers to become more creative, by providing them with both deep and rich insight into their campaigns whilst simultaneously reducing their workload. It is allowing advertisers to be truly “left brain” professionals, injecting innovation into their brands based on hard and fast data about what works and where they have been going wrong.

Predictive modelling is being used by companies like mine that allows advertisers to test hypothetical changes in strategy, based on live data. They will be able to change the course of their approach with little to no risk, having pre-tested the alteration before committing any media buying spend. This fully optimised approach will have an incredible impact on the success of future campaigns and ROI.

From self-correcting space robots and digital dragons, to highly targeted display ads, AI is everywhere. Is this a sci-fi take over? Are robots on the rise? Perhaps, but far from being afraid or hesitant about this, we should embrace the machines as the heralds of the next industrial revolution.