下载PDF

Kyvos Insights > 实例探究 > Viewership Analytics: Leading Telecom Accelerates Time-to-Insights on 150 Billion Viewer Interactions

Viewership Analytics: Leading Telecom Accelerates Time-to-Insights on 150 Billion Viewer Interactions

技术

分析与建模 - 大数据分析
基础设施即服务 (IaaS) - 云计算

适用行业

电信

适用功能

商业运营

服务

云规划/设计/实施服务

挑战

The telecom company was facing several analytical challenges. The time-to-insights was too long for ad hoc analysis, they first had to write, tune, and optimize their queries and then run them to get the required reports. As massive datasets were joined at runtime, queries would take 10 to 30 minutes to return, and complex ones would take even longer. As the business grew, it became difficult to meet the aggressive timelines of its users. In addition, though their data platform allowed them to retain 15 months of data, it was very challenging to analyze across multiple months due to the volume and complexity of their data.

关于客户

The customer is one of the largest telecom and mass media companies in the US. With more than 30 million unique monthly users, they were collecting almost 10 billion records each month. Their data was rich, and they wanted to analyze it across dimensions, such as viewer demographics, network information, program details, and more, for insights to strengthen their business. To improve offerings and grow its business, the company wanted to deepen analytics on its viewership data.

解决方案

To overcome performance and scale limitations, they deployed Kyvos on their AWS platform. Kyvos’ revolutionary Smart OLAP™ technology helped them aggregate 15 months of data on the cloud, and ML-powered Smart Recommendation EngineTM brought in the intelligence required to build optimized OLAP models on complex data. With Kyvos as their BI acceleration platform, analysts can quickly create dashboards and reports without writing any code, and business users can get insights from all the available data. They can slice and dice across multiple dimensions, deep dive into any metric, and still get instant query responses. This was transformational for them as time-to-insight went down considerably. Another key advantage was that Kyvos integrated seamlessly with MicroStrategy while providing its native visualization tool to support multiple business users and teams.

运营影响

Kyvos enabled interactive analytics on 150 billion transactions with the ability to understand viewership by household patterns, time of the day views, programs watched, devices or platforms used, playback mode, and other useful metrics.
Faster and more in-depth insights helped them negotiate better with networks, evaluate the performance of broadcasters and service networks, analyze program performance and compare it with originals, make business decisions based on the origin of content, and identify ways to improve their content, products, and offers.
Kyvos helped them build a scalable, self-service analytical environment on AWS that enabled deep analytics on as much data as needed. Not only could business users respond quickly to network and content changes, but Kyvos also reduced the workloads of analysts and engineering teams.

数量效益

Queries returning in seconds on 15-month data vs. minutes and hours spent earlier, even on single-day data
Reduced coding effort
Optimal utilization of AWS resources with the ability to scale BI up and down to deal with varying loads