Data Drunk
Big data is transforming the World, but there are aspects of humanity
that remain joyfully . . . .  unpredictable


Here’s a data point to end all data points.  Every day last year, Amazon Web Services deployed almost enough server power, to support the company’s entire business in 2005[1].  We are gorging on big data as if it were some form of life affirming elixir; but there is trouble ahead if we really believe that historical data (even, ‘real time’ data is historical) is the sole source of insight.

The fatigue-inducing effects of big data-driven engagement are brilliantly summarized by Spanish academic and researcher, Enrique Dans[2]:  “. . (the use of such) models leads to abuse, to intrusive formats, and above all, to the principle of quantity over quality, which measures success in terms of traffic through websites above all else, including common sense. There are sites — many — that automatically reload their pages every few minutes to add a page view, as well as others that are filled with clickbait. The more page views, the more money . . . .”

Then, there is the thin line between ‘optimisation’ and ‘manipulation’.  UK supermarket chain’s former CEO, Justin King, actually used the term “disloyalty cards” to describe one company’s rewards programme: “We (ourselves) were thinking of them in a disloyal way, working out how to game the customer,” he said[3].  Allegations in the US of so-called ‘false news’ being pushed to subscribers of Facebook and other platforms has further eroded confidence in and credibility of algorithm-optimized content on line.

The impact on brands is palpable; one recent survey[4] suggests that up to a quarter of respondents claim to ignore social posts or ads from brands on social media; many consumers feel bombarded by brands on social platforms, with 34% saying they feel “constantly followed” by online advertising.

Such developments have also helped to raise the profile and value of earned media. According to Procter & Gamble and MSLGroup speaking at PRovoke ’16 earlier this year, when it comes to turning influencer marketing into sales, earned media outperforms paid—often by a factor—of seven or more[5].


Despite evidence of its shortcomings, the collective deliria around big data as a panacea for brands remains.  I recall meeting a ‘Global Web Evangelist’ from one of my previous agencies at a conference and asking him about the viability of a subscription-based platform which promised no data collection or optimization. My thinking being that such a model would, at least, enable consumers to quantify the value of complete privacy against the benefits of a free, but potentially invasive, alternative. No such platform exists; does that represent a gap in the market? My guru/colleague got to his feet, turned and left the room . . . I’d clearly committed Big Data Blasphemy.


The benefits of big data are evident; now and into the future.  However, not at the expense of relationships, instinct, insight, curiosity and humanity.  Sometimes a feel the only sober one at the Big Data Bar!


What algorithm could have predicted, Leicester City would win the 2015/16 Barclays Premiership, or that the male lead from French-made, black and while silent movie (The Artist[6]) would win an Oscar as Best Actor in 2015, or that the next US president would be a former extra from Sex in the City[7]?


If anyone is left in any doubt regarding the shortcomings of big data, and the beautiful, unpredictable spirit of human nature, then show me the algorithm behind this –









Leave a comment