Real-Time Viewership Analytics on 150B+ Interactions

About The Customer

America’s largest broadband and cable operator powering millions of connections

For nearly three decades, this US-based connectivity provider has been delivering broadband, video, mobile and voice services to over 31 million customers across 41 states. Operating under a fiber network spanning nearly 800,000 route miles and supported by a workforce of around 95,000 employees, the company continues to expand its network and invest in advanced technologies to provide fast, reliable and secure connectivity nationwide.

Challenges

Slow analytics and limited data scope restricted audience insights

The organization struggled to analyze massive, fast-growing viewership data, as existing systems couldn’t keep up with billions of records generated each month.

Complex, high-volume data delayed ad-hoc analysis: With 10 billion records generated monthly from 30 million+ users, runtime joins across large tables made queries slow and inconsistent.

Incomplete data visibility: Although 15 months of data was available, analyzing across multiple months was difficult due to the size and complexity of datasets.

Inability to slice and dice across multiple dimensions: Users couldn’t analyze data across key dimensions such as demographics, network information and program details, limiting their ability to uncover deeper audience insights.

Business Goals

Unlock faster, multi-dimensional viewership insights

The company aimed to modernize its analytics stack to better understand audience behavior and support data-driven decisions. Their objectives were to:

Scale analytics for massive datasets.

Reduce query and reporting times for faster business actions.

Enable self-service exploration to allow business users to independently analyze viewership data across multiple dimensions.

Leverage historical data effectively to identify trends, patterns and opportunities for growth.

Strengthen business decision-making by helping teams evaluate broadcaster performance, negotiate better with networks and optimize programming strategies based on audience insights.

How Kyvos helped

Enabled faster, deeper viewership analytics on massive data

Kyvos semantic layer, deployed on their AWS platform, allowed every team to interact with 15 months of viewership data. It built optimized semantic models on complex datasets, overcoming performance and scale limitations.

Unified view of all data: Provided a single version of truth across the organization, enabling analysts, business users and leadership teams to access consistent metrics across all programs, networks and demographic segments.

Optimized resource utilization: Reduced cluster strain, eliminated the need to free up production resources and ensured scalable AWS operations for varying loads.

KPI-driven, self-service analytics: Explored 150 billion transactions across multiple dimensions to track viewership, program performance, device usage and engagement to enable data-driven decisions and content optimization.

Simplified reporting: Allowed analysts to quickly build dashboards and cross-tabular reports. Business users could explore data across multiple dimensions, change filters and drill down to specific metrics.

Seamless tool connectivity: Worked natively with Strategy, supporting multiple business teams simultaneously.