Data and insights are at the center of operating a cable television network. For AMC Networks—a publicly traded company that owns and operates recognized brands such as AMC, BBC America, IFC Films, Sundance TV and WE TV—the stakes are exceptionally high.
“We are trying to optimize every aspect of the business—from marketing campaigns and promotion to scheduling,” explains Vitaly Tsivin, executive vice president of business intelligence for AMC Networks.
Clearly, this requires taking a data-centric approach to the business. “Today, successful businesses are on the lookout for insights made possible by processing large volumes of data, applying machine learning and developing the right statistical algorithms,” Tsivin adds.
For example, when the company markets and advertises its TV shows inside and outside its own network, it’s necessary to use first-party internal data and third-party external data (from both publicly available and private sources) to understand how to approach and optimize campaigns. This includes content it distributes through Netflix and Amazon.
In years past, this meant plugging in data from a mélange of spreadsheets and tapping a variety of business intelligence and analytics tools to gain limited information. More recently, it also meant using specialized appliances that had grown increasingly limited and unwieldy in today’s cloud environment.
However, the existing tools often left business analysts inside the company somewhat in the dark about key factors and unable to focus on key performance indicators (KPIs) to the extent desired. What’s more, those older tools and technologies often devoured substantial staff time and resources.
“We continued to take incremental steps to adopt hybrid solutions and that eventually led to machine learning,” Tsivin says.
Opting for a More Sophisticated Approach to Data
Management realized that the company required a more sophisticated approach to data analysis. In mid-2017, AMC Networks began using IBM’s newly introduced Integrated Analytics System, with its collection of built-in and highly integrated data science tools that work across private, public or hybrid cloud environments.
The components include the IBM Data Science Experience, Apache Spark and the Db2 Warehouse. Together, these and other tools, including IBM Power AI, deliver a collective workspace that allows AMC’s data scientists to generate new analytic models that developers use to build intelligent applications faster. The open-source Apache Spark platform enables faster in-memory data processing by allowing analytics tools to process data where it actually resides, Tsivin explains, rather than having to bring the data to the tools.
The approach of bringing the system to the data has reduced the demand on IT resources, including time and money previously spent connecting systems, managing data and ensuring that APIs functioned correctly. It also has led to better data connections and fewer errors and technical problems, Tsivin reports.
The decision to use the IBM platform came after careful review. “The ability to move from a data solution involving multiple vendors to a single vendor has introduced significant efficiencies and reduced the need for support,” Tsivin adds. “And the ability to use Db2 with in-memory cloud offerings provided a level of richness that wasn’t available elsewhere.”
Part of the decision to move to the IBM platform hinged on a need for ultra-high-performance data analytics, along with flexibility and scalability. “When you look at the market today, you see two types of solutions,” Tsivin observes. “Those are dedicated, mostly local-based appliances and cloud offerings.
“This system relies on a hybrid approach. It wraps a cloud-based Db2 in-memory capability with an appliance. This allows us to manage data locally, but push it into the cloud when we want to share it. It’s essentially the best of both worlds.”
The offering also includes machine learning libraries that further automate and improve analytics processes.
However, the biggest gains have come from increased system performance and deeper business insights. Not only does AMC Networks have better visibility into the marketing and promotion of its shows, it also has gained visibility into day-to-day business needs. These range from determining what types and formats of ads work best in various situations to more strategic issues such as deciding which business opportunities to pursue.
“The environment allows us to dramatically speed data processing and optimize algorithms faster and better,” Tsivin reports. “We now have a lot of powerful capabilities in one place.”
Tsivin says that AMC Networks will delve deeper into the IBM Integrated Analytics System to gauge key performance issues and indicators, and will also look to expand how it uses the analytics and AI framework.
“There is a lot of dark data out there—unstructured data in the form of audio, video and images,” Tsivin points out. “We normally stay away from analyzing this data because it has traditionally demanded enormous processing power and highly sophisticated approaches. But the capabilities and processing power of the Integrated Analytics System translates into new opportunities to put this data to work more effectively.”
Samuel Greengard writes about business and technology for Baseline, CIO Insight and other publications. His most recent book is The Internet of Things (MIT Press, 2015).