Akamai Diversity

The Akamai Blog

Optimize Video Streaming Performance to Reduce Viewer Churn

Why Improve Viewer Experience? Why Now?

As online video viewership continues its march to digital mediums and platforms, viewer expectation around the quality of experience is also maturing. Increased video streaming choices and the low cost of switching providers are leading viewers to churn if the viewing experience is poor.

There is a definite association between the viewer's quality of experience and the video quality, which is directly impacted by the video delivery performance of the CDN involved. Delivering an optimal video delivery performance is one of the important success criteria for CDN vendor selection across the globe.

Image 1.jpg

Framework to Optimize Akamai Video Delivery Performance 

More than three-quarters (77%) of the survey respondents in the February 2020 edition of the Nielsen Total Audience Report marked playback quality to be a very important video streaming attribute. With a focus to deliver an optimal video delivery performance for our media and entertainment customers, Akamai has been investing heavily in initiatives to help customers improve their viewing experience for their end users. 

Working with customers proactively to identify and execute optimization techniques unique to their use case is at the heart of helping customers improve the viewing experience of their end users. We recommend adopting the Video Performance Optimization framework to engage customers in proactively optimizing their video delivery performance based on their specific requirements.

Optimizations Blog, 11.4.jpgVideo Performance Optimization framework

The framework, which is a 5-step iterative process, is meant to help Akamai align with customers to proactively identify and execute optimizations to help them deliver an optimal video delivery performance at a regional or global level. The framework enables Akamai performance experts to use defined steps, processes, tools, and expertise to identify and execute performance optimizations for customers and help them visualize the improvements using accurate CDN performance measurement metrics. 

Identifying performance optimizations, their analysis, and visualization is accomplished using accurate CDN performance data through Quality of Performance (QoP) metrics. These metrics are a proxy to viewer experience and act as a bridge between the video quality and performance measurement gap. Through the adoption of QoP metrics, customers can understand where to implement performance improvements to positively impact the viewer experience.

The next section of this blog describes two case studies where Akamai engaged with customers proactively using the Video Performance Optimization framework to drive positive performance optimization outcomes.

Case Study: Optimizing Performance for a Digital Media Device Manufacturer 

Background:

The customer manufactures a variety of digital media players and devices for video streaming that are used by consumers in North America, Latin America, and parts of Europe. They also operate a popular AVOD TV streaming platform that enables content providers and advertisers to reach global audiences.

The customer launched its ad-supported streaming service on its devices in 2017 and was leveraging competitive solutions for video delivery and multi-CDN load balancing. Akamai's engagement with the customer started at the beginning of 2019, with a small percentage of its streaming traffic in North America.

Problem Statement:

  • As the customer started planning for its international expansion (U.K. in the first half of 2020 and Latin America in late 2020), it decided against establishing an overseas origin in favor of continuing to serve global customers using the domestic U.S. origin infrastructure. Unfortunately, this led to increased server infrastructure load, which impacted performance of the end users. The customer engaged the Akamai team for help in optimizing the end-user experience. Akamai introduced QoP metrics to further investigate the problem in detail.
  • [Step 1]: On the eve of its launch in the U.K. in February 2020, the U.S. origin infrastructure started having issues. Due to lack of redundancy built on the origin infrastructure, there was performance impact for users in the U.S. market.

Solution:

  • The customer started engaging with Akamai deeply to leverage the global scale of the Akamai platform, our differentiated video delivery products, and our expertise to avert the issue.
  • [Step 2]: As the customer was not using any client-side analytics tool to measure viewer experience and wanted to base measurement on accurate CDN performance data, CDN offload metrics and Akamai QoP data were used to access the current performance situation in different regions.
  • [Step 3]: Providing stable service quality at launch without affecting the quality of experience for users in the U.S. was identified as the success criteria for launch after the initial setback. Akamai QoP data was used to measure the stability of the CDN service quality.

Image 3.jpg

  • [Step 4]: Cloud Wrapper was also introduced by the Akamai team to improve the offload and deliver a consistent and stable quality of service.

Image 4.jpg

Case Study: Optimizing Live-Streaming Performance for a U.S. OTT Content Provider

Background:

The customer is an American OTT service featuring full programs as original content as well as content from other partners. The customer provides a diverse library of content, including movies,  TV shows, original programming, sports, and live news.

The customer was using three CDNs (including Akamai) for its live-streaming service, evenly weighted at one-third share for each CDN. In its desire to continually deliver higher quality and improve on the viewer experience, the customer wanted to optimize the video delivery performance before an upcoming major sporting event. 

Problem Statement:

  • [Step 1 and Step 2]: Akamai and the customer engaged in November 2019, and used third-party quality of experience (QoE) data to review rebuffering, video start, and video playback failures across all CDNs.

OptimizationBlog3, 11.4.pngLive-Streaming Performance (July - October 2019), Third-Party QoE Tool

  • [Step 3]: Conversation with the customer presented an opportunity to improve rebuffering and playback failures before the end of 2019 (i.e., within two months).

Solution:

  • [Step 4 and Step 5]: Since the customer's live-streaming configurations were already optimized to its current state, there was little left to do from a product feature perspective. Performance experts and Akamai's professional services team supported the customer to create a list of recommendations and exert due diligence to baseline the data from the third-party QoE tool. Following were some of the recommendations to optimize video delivery performance:
    • Apply XML Optimizations to remove overuse of temporary variables within a delivery configuration to reduce network load and provide incremental performance improvements
    • Quick Retry for route optimization at a request level to reduce video rebuffering
    • Optimizing caching efficiency on HEAD requests to reduce requests from customer origin and enabling object retrieval from most appropriate Akamai server within the same region

Image 5.jpg

Value Confirmation:

Over a 90-day period, the video start failures decreased by 66%, the rebuffering decreased by 50%, and the video playback failures decreased by 30%; as a result of the improved performance, leading up to a major sporting event, Akamai gained 50% traffic share.

Learning:

With large gatherings on hold due to the ongoing COVID-19 pandemic, streaming events are a popular alternative to in-person experiences. Delivering a successful live-streaming event on a global scale relies heavily on proper planning, preparation, monitoring, and support. Utilizing the Video Performance Optimization framework and monitoring using the QoP metrics would help organizations adopt a methodical approach to delivering massive live-streaming events without a glitch.

Performance Optimization: A Methodical Proactive Approach

Every media and entertainment customer is different. Each custom-designs its digital infrastructure setup and has different media workflow architectures and unique end-user geographical footprints. Each of them customizes its streaming technologies based on its unique business requirements. Optimizing video delivery performance calls for a high-touch process that is capable of understanding specific use cases and adopting a strategy based on those requirements. 

The Video Performance Optimization framework helps Akamai engage with customers based on their unique performance issues and establish success criteria that are meaningful for their businesses. The case studies highlighted above are just two examples of use cases where we have driven positive performance outcomes for our customers. 

If you are preparing for the launch of a key business initiative or want to establish a proactive engagement toward your digital transformation for your media and entertainment business, engage your Akamai account teams on a performance optimization journey to improve the viewing experience for your end users and achieve your desired business outcomes. We are happy to help you dive deep into QoP metrics and zero in on the most impactful ways to improve video delivery performance.

 

Additional Helpful Reading:

  • White paper on "Measuring Video Quality and Performance: Best Practices," emphasizing the need for QoP metrics
  • Blog on "Metrics That Matter: Measuring Streaming Performance" to measure user engagement and video delivery performance