Founded in 2004, Sling Media is a wholly owned subsidiary of EchoStar Corporation, a leading provider of multiscreen smart TV solutions for consumers and television service providers. Sling provides the complete award-winning solution of hardware, cloud services and applications that gives consumers the freedom to discover, control and consume their live TV and DVR content anywhere in the world, on their preferred tablet, phone and computing platform. Sling’s patented TV solutions are available directly to consumers or through service providers that have embedded Sling’s capabilities.

When it comes to devices, data is the unsung hero: understanding consumer behavior can lead a company to create a breakout product, a service that addresses needs your customers didn’t know they had, or even disruptive technology that changes industries. After Sling Media rolled out the Slingbox, the company needed to find a way to use the terabytes of data they were capturing about customer behavior. Sling wasn’t just capturing details about which TV shows are popular—its event logs were full of information about how customers used various devices to view programs, in what combination, and when and why they had to do a hard re-set on a device. Insights like that could lead to ideas for revenue-producing offers, new ways of packaging services, or entirely new experiences that could change the way a generation consumes content.

In 2011, Sling realized that to generate those kinds of insights, it needed to take its data analytics to the next level. The company was using a traditional on-premises data center that, while robust, lacked scalability. “At first, we just needed better reporting with more visuals, but then we realized that with a little more flexibility, we could turn routine data into a catalyst for change,” says Dmitry Dimov, Director of Online Services at Sling.

The company considered standing up its own analytics database, but didn’t want to impose limits on the volume of data it could process—and it needed a way to visually explore and express insights. “To grow our database, we had to secure hardware, but we also needed software that would enable us to be very flexible with our data,” Dimov says. “Cloud computing looked like it might be the right tool for the job.”

Sling began using Amazon Redshift as a low-cost alternative to traditional structured databases. “We chose Amazon Redshift because we needed to provide internal customers and partners with reports, metrics, charts and timelines fast,” Dimov says.

Sling maintains an on-premise datacenter with Hadoop clusters, Hive, and HBASE. When events come in, the data is fed through the Hadoop clusters to organize it. Then the data is copied to Amazon Simple Storage Service (Amazon S3) before moving it into a 6-node, 12 TB Amazon Redshift cluster. Given the volume of data expected to come in, the cluster is likely to grow to 16-20 TB by the end of 2014.

Sling extracts details about viewership, feature usage, logs, third-party data, and other information from Amazon Redshift and then sends the data to Amazon Glacier for long-term storage. An Amazon Elastic Compute Cloud (Amazon EC2) instance runs supporting scripts for the process. The company uses Tableau Software for data presentation and visualization, which it discovered on the AWS Marketplace.

Sling also uses AWS to enable its users to connect to devices, stream video, and get recommendations for programs.

AWS has enabled Sling to allow its partners to create and access reports online, without having to wait for hand-curated reports. Even ad hoc queries are easy to accomplish, now that Sling is using AWS. “If someone wants to know how many customers have to do a hard reset on their set-top box, we throw the query into Amazon Redshift, publish a dashboard on Tableau, and within a minute or two, we have the answer,” says Dimov. “That kind of information is very valuable to our partners, but it’s also valuable to our business development team. By using AWS, we can make decisions about new features and offers very quickly and very easily.”

Using AWS has also enabled Sling to build a back-end environment that can support a larger flow of data and reduce analysis time. “With AWS, you push a button and you get 8 TB, just like that. It’s backed up and there’s no operations footprint,” Dimov says. “It’s incredibly cost-effective.”

The company can scale up for events like the Superbowl and the Oscars easily, as well as daily spikes at prime time and over the weekends, without having to spend resources on hardware procurement, configuration, and maintenance.

Because using AWS has made it easier to manipulate data, the company’s approach to analysis is shifting. “Using Amazon Redshift lets us turn on a dime,” Dimov says. “You can take any question and slice the data any way you like: by operating model, by hour of the day, by device, by demographic. Being able to do that so easily really fuels the discovery process.”

To learn more about how AWS can help you with your big data challenges, visit our Big Data details page: