What is Small Data?
Seemingly insignificant behavioural observations can give us insights about human desires and unmet needs. That’s Small Data. This kind of data is the source of some of today’s biggest breakthroughs and brand turnarounds.
Small Data is a subjective approach whereas Big Data is more quantitative. What are the flaws with leveraging only Big Data? Any examples of how companies lost out by relying too much on Big Data?
Companies across the world are convinced that they are on top of what’s going on with Big Data. But nothing could be further from the truth. It has become fashionable to refer to Big Data but unless it is counterbalanced by Small Data, the former is a loose cannon without any direction.
Big Data looks at correlation, Small Data focuses on causation — the reason behind a particular observation. You can’t begin to draw a correlation before first identifying the causation, because it almost always points to a larger context. For example, in 2012, Google concluded that it could predict a flu outbreak days before it would happen based on search terms, and that doctors and pharmacists across the country would be able to order medicines in advance. This was deemed revolutionary. Yet just recently, the Center for Disease Control revealed that the data from Google was twice what it should have been. They found that when one begins searching for terms such as flu, people around follow, resulting in a misinterpretation of the data. The same happened in this case. Like others, Google focused on correlation but forgot about the causation.
In another example, not long ago, a major US bank misinterpreted the increased “churn” at the organisation. Thinking that customers are on the verge of exiting the bank, it prepared letters asking them to reconsider the move. Before mailing the letters, though, the bank executives discovered something surprising. Yes, Big Data had uncovered evidence of churning, but the data didn’t peek into the customers’ lives, and so