How Far Can We Rely On Big Data?

If there is anything Donald Trump’s election to the highest office in the U.S. has taught us, it is that the media can influence elections to go one way even when their ‘intention’ perhaps was to make it go the other way like I explained in a previous article. But then again it has also revealed that in a digital age of marketing our increasing reliance on big data can cause us to fail — sometimes woefully.

Most people were surprised by the election of Mr Trump including members of his own party. Some people who had voted him didn’t think he will win either and neither did I, thousands of miles away on the other side of the Atlantic. Why? Because the polls said he would lose.

For weeks, Mrs Clinton had been leading in the polls and we were all certain she would be the next president of the USA. After much post-election analysis it was revealed that the demographic who turned the table around were the silent majority — those blue collar workers across industrial and rural America who had been forgotten by the polls but are a force to reckon with when they choose to stand up. And they did.

Marketing has always relied on data, but as we move on into the digital age, the volume of data we mine, store and analyse increases. According to one study by EMC Corporation, digital data will grow 1,000 percent between 2013 and 2020 to 44 trillion gigabytes. To put this in perspective, an average family generates enough data to fill about 65 iPhones per year; by 2020, the figure jumps to over 300 per year.

At this point we are faced with the great problem of knowing which data is relevant and which is not.

It is an arduous task that if done well can lead to great results and if done badly can lead to…well…Trump. We marketers are becoming increasingly reliant on big data and can now hardly do a meeting without pulling up past trends and results and base future forecast on them. Not that there is anything wrong with this.

But in the midst of all this, there is a disconnect. We sometimes forget that human emotions does not always follow a trend. We forget the changing cognitive dissonances as our customers move up in life and react to external situations like changes in the economy and their very own psychological and analytical landscape.

We forget external stimuli. We forget human unpredictability and that some random guy is perfectly capable of walking into a store with the intention of purchasing a pillow and ends up buying a spade and lollipop instead.

And just as Trumps’s rhetoric appealed to a group of people who were not accustomed to voting, so too can consumers’ ‘environmental rhetoric’ nullify data and show cracks in analysis.

Human emotion and will will never completely be understood by a machine.

It is thus important for marketers to always go back to the drawing board and not forget that they are dealing with complex beings and knowing that emotional engagement and a deep understanding of the consumer is always the way forward.

Marketers have to define a way to blend data and their own human understanding of their customers to be able to obtain best results. You may argue that artificial intelligence and revolutionary data mining techniques can do this. True.

But Siri and Alexa will never understand a human being like you and me would. Big data, though essential, can only do so much.

Daniel is passionate about marketing and creative writing. He is a champion at “The Big Bang Theory” trivia. An avid fan of “Game of Thrones” and John Grisham, he has been a Manchester United supporter since he was 10. If it’s your thing, you can follow him on Twitter as well.