Does Big Data Require A Big Rethink?

Is it possible that marketers have gotten Big Data wrong -- or applied its tenets incorrectly? Five years into its use, it’s starting to look that way. A growing number of data scientists and information analysts are pushing back, giving Big Data a definitional rethink.

And maybe marketers should too.

To date, Big Data has been defined by the “three Vs” or Volume, Variety, and Velocity. That means the amount of incoming data, source diversity and the speed it accumulates. Or, as a colleague of mine says: “Big Data is any information set beyond your organization’s ability to handle.”

If that’s true, then the key to Big Data is little details -- small data that when amassed in terabyte sizes (numbers with 12 zeros), becomes “big.” So rather than worrying about storing and acting on voluminous data, brands should focus on little details that collectively can make for greater impact.



For many, Big Data’s most obvious usage might be through video surveillance and facial recognition technology. A recent 60 Minutes segment featured a look at how “faceprints” are rapidly becoming the fingerprints of the future, breaking down the visual information contained in a face and translating that data into tiny nuggets. And with a large enough data set. faces are beginning to be paired with users’ Facebook profiles, discerning shopper likes, dislike and long-term desires. But even here, it’s important to appreciate that Big Data is about accumulating highly specific “little details.” 

Three small data steps can yield big marketing potential

For marketers, a simple definition for “small data” might be any information bundle that can be manipulated, analyzed and acted upon in Microsoft Excel. That means databases up to about 1 million rows and 16,000 columns. Of course, there are several CRM/CEM software programs that can tackle similarly moderate data sets. What does 1 million customers look like? That’s the soon-to-be size of UK mobile carrier EE’s 4G network. One million customers may not sound like much, but it is bigger than many think.

Here’s how brands start thinking “small:”

  • Determine which small data metrics are most valuable to your brand. For instance, tracking Facebook "likes" could be meaningless. That is, unless additional resources are available to measure how those "likes" are converted into actual purchases.. While not conforming to the three Vs, the complexity of manipulating data should be an important corollary, as it requires expensive software as well as staff trained in its usage.
  • Once you’ve chosen a data set including types of purchases, purchase location and RFM (recency, frequency and monetary – i.e., the purchase value), the next step is selecting your audience. As a loyalty professional, I recommend starting with your brand’s highest-value customers: loyalty program members -- a smaller, more manageable data set. Measuring loyalty program success and the extent of customer brand engagement is critical in expanding business beyond loyalty members.      
  • Actually begin accumulating data. Quality of data, or data hygiene, is essential. So if your brand intends on measuring specific metrics, then measure them. If the data set your brand is accumulating proves too large or direct ROI is too hard to realize, consider cutting back -- but only after initial measurements have begun. Secure and continue to invite C-level buy-in, so the entire team knows how consumers are being measured and tracked.

The age of mini-measurement is here to stay

Whether it's individuals downloading apps that track the number of steps they have taken per day, or business intelligence software startup SiSense creating visual representations of hard-to-manage data, people are enamored with data collection and the actionable conclusions it can help them attain.

Big Data has not lost its relevance or importance. If anything, it will keep on getting bigger. But for many companies the gem of understanding comes from realizing the big potential in small data sets that, taken together, create a highly granular customer picture -- like mapping digital dots onto a face and deducing shopping preferences. Before marketers expend precious time, money and effort wrapping their heads around truly Big Data with terabytes and exabytes of unstructured information, it’s vital that they start small.

Action these three steps first and your brand will soon be able to handle any information set -- no matter what its size.

Next story loading loading..