Ad Streaming Analytics Case Study

About the Client

They are oldest and longest continuously operating subscription television service in the United States. It airs blockbuster feature films, original series and made-for-cable movies, documentary series, comedy and occasional concert specials, and documentaries. The network operates seven 24-hour, linear multiplex channels as well as a traditional subscription video on demand platform.

Incorporated 50 years ago, The Client is one of the major streaming companies with over 50 million+ subscribers. It has continuously improved its product offerings and services across generations.

Why Choose Infoservices

The absence of a recommendation engine that recommends apt advertisements to right audience at the right time was the major business challenge.
The current state architecture and design did not support streaming and processing real-time data.
Execution challenges included lack of understanding the need, data availability, scalable architecture, and tools to operationalize the insights engine in real-time with high availability.

Technologies Used

Amazon AWS

One of the main reasons for choosing AWS platform is that AWS offers native streaming services that enable the enterprise scale with ease. And the company already had its data lake built on AWS S3, so enabling SNS, SQS, S3 Notifications, and Delta-Lake was easy.

Also, AWS provides unmatched breadth and depth of machine learning, big data, and analytics capabilities, enabling clients to choose the right tools for their marketing technology stack. It removes heavy lifting and complexity, and enables to achieve greater cost-efficiency than any other cloud provider.

Partner solutions

The Info Services team architected the solution by utilizing the best practices and came up with a framework that was scalable, sustainable and performance efficient.
An ingestion framework was built that collected ad sense data from the ad provider and stored it in S3 bucket.
When new files are pushed, S3 events notifications are pushed via SNS and SQS, which then are filtered and stored as delta lake tables pointing to another S3 location. Using delta lake tables, companion delta tables (supporting ad sense files), and snowflake tables are combined to form the final target table.
This data is exposed to data science, analyst, and business intelligence teams for rapid consumption in real-time.

Results & Benefits

The ingestion framework built by Info Services now enable the client’s marketing team to better understand customer behavior and advertisement preferences, thereby helping to devise strategies for more personalized and contextualized advertisements.
At the same time, The Client’s IT teams have an improved and scalable architecture now. And customers have a great user experience watching content with low latency and advertisements they are interested in.