Big Data, Predictive Analytics, Machine Learning Technology, the Internet of Things, Data Science, Unstructured Data, Regression Models, Hadoop Clusters, Data Lakes…

Have I successfully overwhelmed you?

The topic of Big Data can be intimidating. At times, Big Data Platforms can seem like magic boxes, environments that you feed all the data on which you can get your hands and in return, they will explain every aspect of your business. Other times, Big Data is presented as requiring resources highly skilled in applied mathematics or computer science to implement exceptionally complex algorithms. While it is true that having these tools available to you could bring a wealth of knowledge to your organization (two great examples are explored by the New York Times and Bloomberg), how many organizations have self-learning technologies and math PhDs on hand? How can your organization start harnessing the power of Big Data without making the immense investment these technologies require?

As industries have matured, competition has increased. Technological advances have lessened barriers to entry and costs of production, improved communication vehicles have allowed more companies exposure to potential clients, and, especially recently, substantial amounts of information have allowed companies to operate more optimally. In order to keep up with the changing marketplace, companies have turned to Big Data to provide more insightful information about their clients and adapting their decision making to the demographics of their customer base.

This brings us to your organization. In response to the new environment, you have begun to capture new information such as customer information (age, gender, location, etc.), but you are not sure how to utilize this information to help drive the strategic vision of your company.

One way for organizations to tap into the potential of new, diverse data sources and infrastructures capable of processing them is to leverage data to accomplish a familiar objective—to identify opportunities. Instead of looking at Big Data from a top-down approach, try going from the bottom up. Identify a target market that, based on your own business acumen and understanding of your industry, you believe is of high value. This can be traditional customer bases (mothers to be, like in the Target example by the New York Times) or more unconventional, internal groups (employees likely to leave, as in the VMware Case Study by Bloomberg). By first identifying your target group, you can begin to segment your historical data to see what the differentiating qualities are:

• Of the customers that participated in an offer, were many of them from a certain region?
• Of the customers that unsubscribed, how many were under the age of 30?

After you’ve identified some telling factors, you can then broaden the analysis to the entire data set:

• 80% of participants came from within 10 miles of your location, was this specific to when the deal was running or is it generally the case?
• Half of all customers that unsubscribed were under 30, how does that match up with the ones that stayed?

If it looks like you have discovered a data anomaly, you are well on your way to having actionable findings! Presuming you have a large enough dataset, your initial findings could be directionally correct enough to be worth additional investment in the group. Perhaps you concentrate more of your marketing efforts for an offer in certain areas, or you structure a new subscription package for customers under 30 whose contracts are ending.

If you want to refine your target market even more, you can bring in more customer qualities and see if there is significant information gained by adding additional criteria to your data segmentation. You can get a better look at the math here, but the cornerstone of this idea is adding additional criteria (home owners under 40 vs. simply home owners) will give you an even better sense as to which groups should be targeted.

While this method will not provide the pinpoint precision that self-learning technologies or mathematicians have the potential to provide, it does provide valuable insights and has the potential to unearth substantial opportunities while minimizing barriers to entry. Perhaps more importantly, this exercise will begin to train your organization in how to better navigate and understand your data. This will allow you to start taking strides on your path to Big Data maturity, potentially leading to the more advanced practices more prevalent in popular case studies. Until then, start small, build some momentum, and start reaping the benefits of Big Data!