Correcting Five Myths About Big Data

We really care about data-driven marketers here at the DMA, and we want to help them succeed.  One of the biggest buzzwords in the industry right now is “Big Data,” which has become a sort of catch-all term for the vast amount of information generated by our digital lifestyles, and the analysis techniques for dealing with it all.    That’s why we were so interested in this column by Samual Arbesman in the Washington Post, titled “Five Myths About Big Data.”  Since big data is extremely important to businesses, but poorly understood, it’s important to take a step back and separate truth from hype.  Here are the five myths he identifies:

1. “Big data” has a universally accepted, clear definition.

Not so!  Lots of people have trouble with what criteria to use in defining “big data.”  It’s become a pretty vague term — and that makes it easy to use in all kinds of contexts — including contexts where another term might be more appropriate.  Just because an application is useful and involves the use of data doesn’t necessarily make it “Big Data.”

2. Big data is new.

While the sheer volume of data available in this day and age, and our ability to process it at a high level are certainly new, the concept of correlating and analyzing vast volumes of information is certainly not.  To name just one example, huge cross-references of every single word used in the Bible, called “concordances,” were in use by scholar-monks for centuries.

3. Big data is revolutionary.

In his Post column, Arbesman compares the effects of Big Data’s rising prominence to the effects of the Gutenberg printing press: Revolutionary at the social and economic level, yes, but less so at the individual level for a while.  The impact will be modest and gradual for the average person, he said.

4. Bigger data is better.

As with many things in life, bigger doesn’t necessarily mean better.  That’s why it’s important to have experienced personnel managing your data.  Those managers should be able to judge when a huge number-crunch is called for to solve a problem, or whether a simpler type of analysis of a more limited dataset will be sufficient to your needs.  Data on its own can’t tell us anything—we need people to interpret it!

5. Big data means the end of scientific theories.

This might seem an outlandish claim to begin with, so it’s not surprising to us to find Arbesman saying it isn’t true.  He  refers to a well-known claim from Chris Anderson in a 2008 Wired essay which stated, essentially, that if enough raw information were gathered together and sorted, correlated, and/or analyzed, we’d be able to pull associations and relationships out of any pool of data without actually having to observe, test, or experiment.  But that doesn’t allow for a deeper understanding of causation (which, as we all know, is not the same as correlation), nor does it allow for the amazing people who can take data and correlations, and question their impact and implications.

As important as big data is, it will never be more valuable than sharp, inquisitive people.  So celebrate yourself and every data-driven marketer you know, and keep looking for ways to make big data a useful part of your business without overstating its impact.

##  ##  ##

This post is based on an article in The Washington Post by Samuel Arbesman, an applied mathematician, network scientist, senior scholar at the Ewing Marion Kauffman Foundation, and the author of “The Half-Life of Facts.” Follow him on Twitter: @Arbesman.

Your Comment