The author frames the challenge succinctly (and correctly).  Granted it is more B2C oriented than B2B, but the B2B marketer will understand the impact (and challenge):

Modern business social media monitoring and analysis demands a disciplined, scientific approach, rather than a non-scalable, miscategorised overview of online opinions. Companies must systematically implement a scientific methodology that moves from mere social listening to social intelligence. This means analysing the content of Big Data, respectfully extracting meaning from public conversations, and making actionable sense out of what is being said. Only in this way can social sources be used smartly to activate and recalibrate marketing, customer service, competitive insights, fruitful prospecting and new business programmes.

Of course it’s also important to refine scientific analysis by bringing human judgement into every equation, incorporating a ‘teachable’ algorithm that finds real meaning behind Big Data. It’s critical not to misinterpret artificial results that might skew business intelligence and result in bad decisions.