Our Blog

Using data to deliver advertising before a need emerges

Julian Assange took a typically reserved stance on the subject of privacy recently. “The ability to surveil everyone on the planet is almost there,” he decreed at SXSWi from his Ecuadorian bunker in the heart of Knightsbridge.

To be honest, when it comes to filling the news pages, the integrity of personal data didn’t need the world’s most famous fugitive to raise its profile. In fact, if a recent conversation I had is anything to go by then it’s sure to remain a distinctive trend for a little while yet.

A friend, who recently left the military and found employment as a government consultant, was telling me about a project he’s working on to build a single holistic digital view of every citizen. The theory is that if you can combine demographic information with different behaviour patterns and apply some clever statistical analysis, then you can predict how that individual is likely to behave in the future. This has obvious benefits across a number of sectors like tax and health, but I suspect given the type of people involved it had more to do with security. Still, it’s good to know that our industry is wrestling with the same challenge as GCHQ albeit with slightly different objectives.

The Holy Grail of holistic marketing

Getting a holistic view of how individuals behave, and then using this to predict how they are likely to behave in the future, has become a priority for brands. Google’s recent £242m purchase of DeepMind is a great symbol of this. Laying such big numbers on a cutting edge artificial intelligence business that has yet to produce anything commercial may initially look the folly of the rich, but it provides an indication of how important predicting the way people think will become.

Amid the hype, theoretical chat and grand ambitions, there remains a significant gap between what is possible in the future and what we can do in the short term. There is a real need to get the basics right first and starting with a ‘stimulus response’ approach in the short term, provides a useful base to build from. For years we have known how broad environment factors trigger group behaviour – if it’s sunny people buy more beer, if it’s cold we buy more tomato soup – but this has hidden individual nuances.

Last year saw some campaigns that started to bridge the gap. For example, our Stella Artois Cidre campaign identified the impact of local weather on purchase behaviour. We discovered that sunny conditions combined with an increase of 2°C on the average temperature in that town triggers an increase in cider drinking. The next step is closing the gap to the individual. If it’s hot, one person may buy Stella Artois Cidre, someone else will purchase Budweiser and a third person could go for a Diet Coke. Getting the right messaging to the right individual, based on how they have responded under certain conditions will be a powerful asset for brands.

A truly holistic view of the individual opens up the potential for a ‘personalised web’ that utilises individuals’ behaviour to determine a need for a product, or can help a brand tailor itself accordingly.


A personal yet predictable journey

Imagine a scenario where a London businessman visits his family in Birmingham on the last weekend of every month. He could be searching for cinema times a few days before his trip, at which point he’s served film times according to his anticipated visit, instead of his current geo-location.

Reports into this kind of targeting are common and will increase as time goes on. I like this example about Google Now posted on Reddit:

“Today, I was heading out for a job interview, and about 20 minutes before I left, I looked at the directions to the place on Google Maps on my computer. I got in my car, opened the Google app, ready to tell it to take me to there, but the card at the top of the page was already asking if I wanted to navigate there.

It knew what I’d just searched for on Maps and applied it to Google Now!” [i]

There have been complaints about the accuracy of Google Now and this kind of ‘mind-reading’ technology is in its infancy. However, it isn’t a stretch to imagine a day when it will be able to tell you something before you asked and perhaps, before you even realised you needed to know it.

For any marketer, building a complete picture of the consumer will soon be vital. Knowing that I clicked on a particular car product page isn’t the same as knowing what I do for a living, whether I have kids or not, if I’m vegetarian or vegan and that I clicked on a given car product page. Facebook’s new product is a real step forward in this space, bringing together a wealth of publicly available information such as when my home insurance is due, how long I have lived somewhere, how old my car is and links this to personicx clusters and other social information.

Whether it is brands, the NSA, or the GCHQ; all are in search of the Holy Grail where the internet and digital world penetrates physical society. The main barrier to predictive holistic marketing will be the extent that online privacy lobbies have their way, but the key to unlocking this will be how we use the information to speak to the individual and create real value for them.

Ian Edwards is head of strategy at Vizeum
This article was originally published on The Wall


Tags: , ,