When talking about Big Data, the first question people often ask me is how I would define it. Though this usually leads into conversations about the amount of data companies are collecting, I believe it misses the true meaning behind Big Data. Big Data isn’t about how much data a company has, it’s about what business value companies can gain from it. The ability to mine data, combine it with other data sources, and analyze the data is what Big Data means to me.
I like to use the real life example of Target to illustrate what I mean. In February 2012, the New York Times published an article on Target about how they used their predictive analytics to market maternity items to a teenager even before her family knew about the pregnancy. How was Target able to make such an accurate and pinpointed prediction? By combining large amounts of diverse data on shopping trends with analysis by highly specialized data scientists, Target was able to change what they marketed to the teen based on the changes in her purchasing behavior. Without the properly skilled analyst team to explore their Big Data, Target would just have a lot of useless information.
Another, more familiar, example of a company that leverages their vast amounts of data is Amazon. Amazon collects massive amounts of data on the browsing, shopping, and purchasing habits of all it’s customers. Additionally, the Kindle Fire’s Silk browser funnels all the user’s web usage through Amazon’s EC2 cloud, providing Amazon with even more data. All this data is used to predict purchasing trends and drive their marketing efforts and recommendations, which in turn increase customer purchases, leading to bigger profits.
These examples prove to me that the ability to gather and analyze data is what defines Big Data and makes it interesting. Having the ability to collect and store more data is certainly very important, but without the ability to mine and analyze that data, it’s pretty useless. The real lesson is that before a company spends time and money on building out a hadoop cluster to store large quantities of data, the company should evaluate its ability to do analysis on data and put the correct resources in place to ensure that the data collected can be leveraged in a useful way.
Brian Rowe is the founder and CEO of Perceivant, a Big Data Business Intelligence platform. Its fully hosted cloud solution helps companies get started with Big Data quickly with low risk.