Case Study
Using Self Service Analytics for Insight driven Decisions and Targeted Marketing for a US Media Company
Overview
The U.S. E&M revenues is expected to reach $759 billion by 2021, up from $635 billion in 2016, increasing at a CAGR of 3.6 percent – holding steady at the same CAGR as last year. - PWC
E & M Outlook forecast for the U.S Industry
The current Music & Entertainment (M&E) value chain has changed, wherein the top-line growth from mature segments is gradually declining and the digital M & E content is growing, though at a lower rate. The primary reason is that the fragmented audience today are offered various options in selecting a multitude of television packages, as per their needs. This has led to fiercer competitions among key M & E players, in gaining much of the overall digital wallet share.

The client, a U.S multimedia conglomerate, currently faces a huge explosion in data volume and variety, with a growing network of clients and digital platforms being leveraged. The system currently has an overhead of handling a data size of about 5+TB, with a data inflow of approximately 65 GB/day. It is now on the lookout for ways to monitor the consumption of its digital content, decide on investing its advertisement returns and further boost up its viewership on selective shows.

The Challenge

While achieving its business needs, the client faced the following challenges:

  • The lack of a single unified platform. The presence of separate Redshift data warehouses in two different locations, one in North Virginia, primarily for ETL integration using Talend, a cloud based Big Data Integration tool, and the other based out of Oregon, for Data Science and Reporting, has increased the Cross-Region Network Charges and delayed the overall Data Processing Speed.
  • Delayed and Incomplete Business Insights: The client’s existing reporting and visualisation options to display viewership patterns and trends are ad-hoc/manual, time consuming and error-prone, and this hindered instinct driven decisions. This also undermined the client’s efforts in increasing viewership and staying competitive in the media industry. The lack of integrated insights has also limited the ability of key business leaders in making true data-driven strategic business decisions.

Our Solution

To address the above challenges,

  • An elastically scalable analytics platform was built on top of the AWS Cloud with the following key features:
    • A centralised data lake system based in one region (Oregon) was used as a single source of truth, to store incoming business data from disparate sources: The viewership data obtained from these sources was analysed and a customised data lake was created using Amazon S3, a scalable and flexible data lake solution. This data was stored in compressed file formats using Parquet in S3 data lake and Redshift, a cloud-based data-warehouse solution. The architecture has used Redshift Spectrum, a key feature to conduct fast and complex analysis on objects stored in the AWS cloud, to offer faster data processing and integration at reduced costs.
    • A platform to unearth business insights from raw semi and unstructured data: Using AWS Redshift, a Massively Parallel Processing (MPP) data warehouse solution with auto-scaling options, accelerated insights could be obtained from structured data. This was further pre-processed and loaded into Redshift, to reduce IT dependency on analytics reports.
  • Talend and DOMO, both cloud-based tools for reporting and visualization support, could offer timely business insights for key decision makers. The data was loaded on OTT (Over the Top) Media services, social impressions and user behaviour data to data lake and analytics platform, to fulfil the following purposes:
    • Answer user questions such as count of viewers who have transitioned in the number of full episode viewers (FEP) among digital platforms
    • Provide a holistic view of business insights for decision makers.
    • Shift from instincts-driven to insights-driven decision making for internal teams (Marketing, Product, Production, Social Media Listening etc.).
Results and Benefits
  • 4x faster data integration: The proposed Re-Host strategy (which uses custom scripts) has increased data integration and processing speed by over 200%, translating into $50,000/year in cost savings.
  • Seam less 360-degree view: With auto-processed data available in the lowest grain from all data sources; slicing and dicing it to unearth insights have been made simple for the internal teams.
  • Quicker Analytics Support: The data available in both S3 and Redshift is now leveraged downstream using Domo scalable advanced predictive analytics.
  • Agile Data platform: The Data Lake platform which stores business data in Parquet file formats, now provides flexibility and portability to move from between visualization and analytics platforms, with minimal changes.