Big Data Isn’t Just for the Big Companies
As Cablefax has previous reported, cable MSOs that can harness “big data” are able to more effectively identify the product and services that particular customers are most likely to buy. They also know the optimal sales price as well as the best time and method for making the offer.
I understand that big data sounds like big bucks and something that’s well beyond most organizational capabilities and budgets. But Conrad Chase, director of big data for media at Data Realty, a provider of hosted big data infrastructure and analytics, says media companies could be missing out on a huge opportunity by not challenging those assumptions.
In an article entitled “Big Data Plunge” appearing in the May-June issue of MFM’s member magazine The Financial Manager (TFM), Chase says “the good news for the mid-market is the big data trend is seeping down into the ranks of smaller enterprises as the tools, technology and talent are becoming more available and affordable.”
He goes on to point out that, “While the adoption of big data by larger entities could be viewed as a threat to midmarket media companies, in reality, data is a great equalizer and is agnostic to company size, brand names and media personalities.”
So how does a mid-sized company start to take advantage of that opportunity to match their products and pricing savvy up against a lager competitor’s?
Chase says most organizations’ first step into big data is a pilot project. It’s a great place for mid-market media companies to begin as well. These pilot projects typically focus on one specific issue, such as, “How can I hold onto my best customers?”
Answering that question will require creating the right physical and cultural infrastructure to allow the company to make the best decisions. “Within media organizations, data is spread out across multiple databases and stored in different formats, which are all critical for understanding the company’s relationship with its users and customers.”
As an example, Chase points out that it’s not uncommon for accounting departments to use one database for billing, while their sales teams have a customer relationship management database for prospecting, with their marketing departments relying on still another data set. Chances are these same organizations are also relying on additional data systems located off-site with multiple vendors.
In order to integrate the “high volume, velocity and variety” of all this information for the pilot project, companies will often turn to solutions like Apache Hadoop, a free, open-source framework written in Java. According to Chase, Comcast Corp. and Entravision Communications are among the media entities that have used Hadoop to integrate data from their call centers, billing relationships, social media interactions, Web site click-streams, bandwidth usage and viewing preferences.
The next phase of the pilot project will involve identifying and integrating external data sets that can provide a more complete answer. These sources include freely available information, such as U.S. Census data, and premium information that could be worth the purchase price; among these are financial and employment data from sources such as Experian and Equifax.
Chase identifies the third stage of the project as using predictive analytics tools that study historical patterns and trends in behavior in order to make accurate predictions about the future. “These results produce a ‘propensity score’ that ranks every individual in a given group based on the statistical likelihood that they will or will not behave a certain way.”
Giving an example of a pilot project focused on how to retain the most valuable customers, Chase said the first stage established a profile of the traits for these customers while the second stage used external data sets to paint a more complete picture. In the third stage, he says the big question can shift to: “Who are the individuals in my target geographic area that are not customers, yet match the profile and traits of my most valuable customers?”
“The answer to this question can create a highly targeted lead list for the sales and marketing teams,” says Chase. “Imagine what a business can do if it knows who has the highest propensity to become the most valuable customers.”
The current issue of TFM is available on MFM’s Web site, www.mediafinance.org. As the saying reminds us, “knowledge is power.” Knowing how to harness data is one way to become more successful in your markets. Even if you don’t do it, I can guarantee that your competition is or will soon.