Wired Magazine’s editor-in-chief, Chris Anderson opened day 3 of DMA 2012 with his take on everyone’s favourite buzzword – Big Data. Chris’s approach to the problem is an imperfect one, best summarised by his quote from George Box on models – “all models are wrong, but some are useful“.
Chris worked through the theory that in the age of the Petabyte, when throwing enough data at a question, the patterns that emerge are far more powerful than theory. He gave the example of Google Translate which, without any pre-concieved knowledge, converts from any language to any other based on user input and overriding patterns – even English to Klingon!
He proceeded to expand on this thinking by exploring the fact that internet has room for everything – the mass to the niche, the large and the small. While traditional business models focus solely on the averages and mass market (think bell curves), the internet provides the opportunity to market to the long tail. That is, those specific preferences or cross-sections which fall outside the mass volume of whichever metric a business records its market against.
This concept alone has (or should have) massive influence on the way marketers approach engaging with their customers. As companies explore the opportunity to reach their customers’ long tail, will the emergence of Big Data and the ability to leverage this market change its very definition?
Chris further ran with this concept into comparing true data-driven, test obsessed companies with visionaries and innovators. He compared Amazon, who as king of the digital A/B test drives its strategy and marketing almost wholly by data, to Apple at the other end of the spectrum. Anderson’s view is that neither approach is right or wrong – without data, accurate decisions could never be made, where without innovation, new concepts would never be born. A dichotomy which is, or should be, close to every modern marketer’s heart.
The age of the geek has arrived and those who make the most of complex, big data wins. The perfection of traditional direct marketing will cease to exist (or at the very least, take a shadow role).
As Chris more poetically put it, correlation is not causation, but it is something.