Customer Care: Taking Advantage of Big Data

Operators Are Learning How Best to Exploit Information Assets

Cable operators have been gathering data on their customers for decades, but that process was often tedious, expensive and manpower-intensive. It also put distributors behind the curve.

By the time the data was collated and analyzed, it was often too late to fix problems or take advantage of opportunities.

With the advent of “big data” techniques, which allow for cost-effective, real-time capture of data, distributors are now able to act and react more quickly and effectively.

The catch: Determining what’s important, what’s actionable and how to get a return on investment.


At this point, most operators are capturing Terabytes of data about their customers every day, according to experts. But many MSOs are still struggling to take that information to the next step with strategic plans and actions.

“The key is prioritizing our goals and then turning those priorities into business-specific outcomes,” Tom Karinshak, Comcast’s senior vice president of customer experience, said. “We don’t want to be overwhelmed by the amount of data we are collecting so clear and deliberate plans and objectives are necessary for success.”

Until recently, collecting data meant putting information into structured boxes and columns, Ryan Pellet, senior vice president of strategy and services for vendor Nexidia, said.

When cloud storage was introduced in 2008, the cost to capture and store data plummeted. The introduction of Hadoop, a free, Java-based programming framework, also enabled companies to better process the large amounts of data they began storing. The age of big data was born.

The amount of information being collected today can be mind-boggling. One unnamed MSO collects about 70,000 hours of audio each day, and the value of that data is just coming to the surface, Pellet said.

A recent study from Cisco Systems, submitted as a technical paper at May’s Cable Show, determined that every 1 million cable subscribers generate 2,400 Gigabytes (GB) of data per day, via switches. Thus a cable operator with 25 million subscribers, or a little bigger than Comcast, would generate 87,600 GB (the equivalent of 87.6 Terabytes) of data per year, just through the switch.

Theoretically, at least, operators can take all that information they have been collecting for years and turn it into real-time solutions and opportunities.

The key is analytics, both Pellet and Yossi Zohar, marketing director of Amdocs’ customer management division, said.

To date, a lot of money has been on the foundation necessary to collect and store all that data, but not much is being spent to analyze it, Zohar said.

And analysis is only part of the key to success. “Once you have clear objectives, the data can help you fine-tune your strategies,” Comcast’s Karinshak said.

Analytics can turn conversations into data that operators can then use to do more than just track connects, disconnects and churn, Tapan Dandnaik, senior vice president of customer service and financial operations at Mediacom Communications, said. They can also help MSOs track where those activities are occurring — and why they’re happening — in real time.

Strategic planning, execution and specific goals are what make collecting big data a good idea that makes sound financial sense.

The ability to translate conversations into data, then turn that data into tangible information, is huge, Pellet said. It helps operators shift from playing defense — that is, trying to fix a problem after it occurs — to offense, or anticipating a problem and fixing it before it becomes widespread.

Data can come from a variety of sources, including online and mobile transactions as well as social-media outlets. But phone calls remain the most popular way for customers to interact with their service provider, and thus the richest data source, Pellet said.

Capturing call data is imperative to a successful approach to proactive customer service. But information gleaned from calls must be parsed and strained in order to find the right information necessary to meet objectives.

Amdocs offers a new suite of products, called CES 9, designed to provide distributors with the ability to provide real-time notifications and recommendations, including issue resolution, discounts, new feature availability and billing notices, according to Zohar. The results can be as much as a 20% reduction in calls to service centers, he said.


Using speech and text analytics, Nexidia products can detect data patterns that can tell an operator which customers are most likely to leave; which customers are having issues; what competitors are claiming; and can gauge new product performance from a customer’s standpoint.

Operators can monitor specific metrics, including average call-handling time, non-talk time and call volume by category. They can also group calls into key subject areas based on phrases associated with a particular call type, Pellet said.

Big data provides Comcast with a “holistic view of the customer’s experience,” Karinshak said. “We collect an incredible amount of data and insight from our customers every day,” he said. “We want to make sure we keep it manageable and so we must stay laser-focused on prioritizing that data and tying it to business outcomes and goals. Once you have clear objectives, the data will help you fine-tune your strategies and ensure success.”