Big Data: let’s bin generalities in favour of specifics

Does anyone else get a little jaded by the levels of Big Data evangelism at the moment?

Too many times I’ve heard excitable Big Data presenters who, when probed, mutter something vague about social media and sentiment analysis. Too many times I’ve heard people describe their Big Data solution as the “next big thing” but are then forced to admit that what they’re offering is little more than what Tesco have been doing with Clubcard for years.

A couple of weeks ago, I came across this graphic from “Big Data Borat”. 

I’ve got it printed out and pinned above my desk. Why? Homer Simpson sums it up best: “It’s funny cos it’s true.” I exaggerate, of course. But specific examples of how to use Big Data, or indeed social media of any sort, are far more powerful than broad sweeping statements. Now Big Data is not all hype – there are some lovely case studies (I’m constantly in awe of some Ipsos colleagues of mine who work using mobile Big Data – to solve specific problems).

I only caught a few minutes of last night’s Horizon Big Data special but fortunately they did concentrate on some of the best examples. The predictive policing in California  is something I’ve had a personal interest in for months – and is certainly worth further reading. On the financial side, they looked at Winton Capital’s efforts to harness analytics to enhance their trading.

I had half expected them to talk about the “Twitter predicts the stock market” model, which gained mainstream media attention. The principle behind Derwent model was fascinating, although not without its critics  (you’ll need better maths than me to understand it fully, but the overall conclusion is clear). Derwent Capital Markets have since abandoned their strategy, although they now provide a social media dashboard service so that independent traders can follow in their footsteps.

It’s easy to be cynical, but I would applaud these intrepid efforts. Social media’s predictive elements may not be quite as powerful as we had initially hoped, but it still provides immensely rich and colourful data with which to guide business decisions…if handled carefully.

A mantra of mine when it comes to social media listening is “specific, not general”. Think about what you’re trying to find out and focus on trying to find out what people are saying that’s directly relevant to that. You could look at what people say online about changing from one service provider to another, or why they don’t enjoy a particular product, or – the holy grail, perhaps – what they wish someone would invent. It’s fantastic that Big Data, and social media data in particular, are getting so much attention. Solving specific problems is where we’re at now.

And maybe, in time, we’ll prove Big Data Borat wrong and save the world.

Eoghan O’Neill is social media listening analyst at Ipsos MORI. He tweets at EoghanLondon and blogs occasionally.