Why Big Data Isn’t a Panacea

As people increasingly use digital communication channels to access information, execute business transactions, and interact via social networks, the volume of data regarding these activities grows exponentially. This massive and growing volume of information is now being called big data, and few topics have received more attention in marketing and technology circles over the past couple of years. According to many pundits, big data can dramatically improve the quality of business decision-making generally and the quality and effectiveness of marketing in particular. It can enable companies to develop valuable insights about what current and prospective customers want, what competitors are doing, and how markets are changing.

The hype surrounding big data has been huge, and many prominent voices have been effusive in describing big data’s potential benefits. Recently, though, several articles and blog posts have attempted to provide a more balanced view of big data. The authors don’t dispute the importance of big data or the value of using data to support business and marketing decisions. However, they do identify some of the reasons that big data isn’t likely to be the “silver bullet” that the hype suggests.

Here are brief summaries of two of these recent commentaries.

The Big Data Fallacy And Why We Need To Collect Even Bigger Data (Dr. Michael Wu, Principal Scientist of Analytics at Lithium)—In this blog post, Dr. Wu begins by stating that data is only as valuable as the information and insights we can extract from it. He then argues that data and information are not synonymous and, more importantly, that more data doesn’t produce proportionately more information because of the redundancy that exists in nearly all data sets. Dr. Wu also argues that not all information will provide insights. He contends that a substantial amount of the information in most data sets is not interpretable and therefore cannot produce insights. And, of the information that is interpretable, some will be irrelevant noise that cannot support valuable insights.

Dr. Wu summarizes his view of big data in these terms: “The value of big data is hugely exaggerated, because insight (the most valuable aspect of big data) is typically a few orders of magnitude less than the extractable information, which is again several orders of magnitude smaller than the sheer volume of your big data. I’m not saying that big data is not valuable, it’s just overrated, because even with big data, the probability of finding valuable insights from it will still be abysmally tiny.”

The big data bubble in marketing—but a bigger future (Scott Brinker, co-founder and CTO of ion interactive, and author of the Chief Marketing Technologist Blog)—Like several other commentators, Brinker believes that big data has been over-hyped. He writes, “First, the expectations of what big data will deliver on its own, especially in the short-term, are massively overblown. Lots of people are throwing lots of money at the promise that big data will, somehow—I don’t know how exactly, it’s technical and mathy—tame the fractured, fragmented, frenzied landscape that is modern marketing and crunch all of those complications into more customers. All as a well-oiled machine. Regrettably, it’s not that simple.”

Brinker contends that the real value of big data is that it can enable a company to implement data-driven marketing. In Brinker’s view, there are three steps in this process.

  • Collect and organize data to extract insights from it. This is the role that big data and big data analytics play in the process. However, the primary output of this data analytics is not firm conclusions, but hypotheses about possible cause-and-effect relationships.
  • Test these hypotheses to prove (or not) cause and effect.
  • Use the proven cause-and-effect relationships to deliver better customer experiences.

 

More Insights

For more insights regarding the value and limitations of big data, take a look at the following:

 

Tags: , , , , ,

No comments yet.

Leave a Reply