By Paul Pellman
Big Data has become somewhat of a ubiquitous topic. Businesses, academics and even those in the political sphere are paying close attention and looking to understand and glean knowledge from this new world in which seemingly everything is measureable.
More than ever before, data insights are poised to provide a demonstrable competitive advantage for those able to effectively analyze it and act on it. What have we learned about Big Data so far?
1) Data Has Forever Changed the Political Process
You’ve probably read about FiveThirtyEight and Nate Silver’s adept election predictions formed by applying statistical analysis to polling data. Political campaigns themselves are actively seeking ways to improve their data warehouses and techniques for utilizing the social networks of supporters to more effectively target key demographics in swing states. This isn’t the last data arms race we’ll hear about, as both the public and private sector attempt to expand and organize structured and unstructured data in order to more efficiently track and tailor how they reach their respective constituencies and customers. Many advertisers are well on their way down this path and are seeing double-digit ROI as a result.
2) The Data Deluge Necessitates New Skills and a Different Approach
Having data is only a small part of the equation however, and more than ever organizations are struggling to make sense of the customer data available to them. Companies such as Facebook are now generating more than 500 terabytes of data on their own every day. As data sets continue to proliferate and grow in size, marketers and IT professionals are increasingly going to need to develop different skills and attract new talent capable of analyzing and taking action based on what the data is telling them. This will be easier said than done though, as McKinsey predicts more than 1.5 million workers with “deep analytical” expertise will be in demand in the U.S. alone.
3) Black Friday Shopping Data Sends Misleading Message
One of the unfortunate side effects of the big data deluge is that, in a vacuum, data can be misleading. Even the most sophisticated measurement and analytics platforms require some level of human analysis and interpretation.
For an example of this, look no further than the data concerning Black Friday referrals, or the lack thereof, from social networks. As attribution professionals, the results were somewhat surprising and contrary to what our clients experienced, likely resulting from IBM’s findings failing to properly account for the role social played in influencing the purchase decision or driving customers to the various channels (search, display, etc.) that ultimately led to a customer conversion. Regardless of individual interpretations, the fact remains that measuring and connecting data points is only one part of the journey. Marketers who are correctly analyzing and inferring actionable insights are those that will be leading their businesses forward in 2013.
4) Effectively Combining Data Sets Can Produce Extraordinary Results
Of course, businesses weren’t the only institutions to embrace structured and unstructured data to make more informed decisions. In 2012, government agencies and academics drastically accelerated efforts to use massive, complex data sets to identify solutions to large-scale challenges.
For example, researchers and authorities monitoring the trajectory of Hurricane Sandy were able to identify, with a great degree of accuracy, which areas along the eastern seaboard were most likely to be impacted by its devastating force. Using these models along with data from 11 previous hurricanes, researchers also projected how many households would be without power (Google later used these insights to quickly supply helpful information on its mapping platform such as the location of shelters and traffic conditions).
The ability to more rapidly respond and intervene when necessary based on near real-time modeling is an important step forward for agencies, researchers and businesses alike as organizations combine massive data sets and attempt to apply machine learning and analytics to an ever-expanding list of disciplines.
Paul Pellman is CEO of Adometry.
Network
