AWS for M&E Blog

NBCUniversal International’s reality on-demand OTT platform hayu adds personalization and data analysis with AWS

Thanks to hayu – a unique multichannel OTT video service that’s dedicated to reality TV programming – international viewers have access to new episodes of their favorite programs on the same day as their US broadcast. First launched in the UK, Ireland, and Australia in 2016, the hayu service is now available in 27 countries across a full variety of mobile, tablet, and TV devices, as well as the web. The service, from NBCUniversal International, provides more than 8,000 episodes from more than 300 of the best reality shows.

Over the past 18 months the service has been upgraded – working closely with European-based media software solutions developer Merapar Technologies – to include new navigation functionality and dynamic content presentation, providing viewers of the OTT platform with more of the content they love. The business benefits are significant, as hayu has boosted the performance, viewer retention, and subscription revenue of its service by deploying virtualized media processing tools from Amazon Web Services (AWS).

Among the full breadth of cloud-based AWS services employed are: Amazon Personalize, which allows customers to create real-time personalized user experiences faster at scale; Amazon Kinesis, a real-time data streaming service that can continuously capture gigabytes of data per second; Amazon Rekognition, an image and video analysis service that can identify well-known celebrities in a wide range of categories; and Amazon Kinesis Data Firehose, a capture, transform, and load streaming data service found in the AWS Management Console.

Merapar received a request from the hayu team to build a custom multi-layered content management solution that adds viewer personalization and user data capture capabilities. The OTT platform generates several billion data events per month that now flow through Amazon Kinesis.

Amazon Kinesis is used to collect, process, and analyze real-time, streaming data to give hayu timely user insights that the team can then react too quickly. hayu also makes extensive use of Amazon Simple Storage Service (Amazon S3) to house their data lake.

“The benefits of using AWS include security, scalability, and reduced maintenance, and these factors were so critical to building the solution that many content providers are seeing AWS as the solution,” said Chris Birkinshaw, technology principal and solutions architect at Merapar and the project lead. “We have been able to move to a different way of developing new features for the platform. We can experiment, spin up, and spin down services as required, without having to worry about the cost.”

hayu is now using Amazon Personalize to provide a better customer experience via tailored recommendations that leverage machine learning algorithms, making it easy to create individualized title suggestions for subscribers of the service. This customization is based on viewing history.

The new platform – for which Merapar developed and delivered a prototype “serverless” content management systems in a mere 21 days, and a finished product in 20 weeks – is a huge leap forward for hayu.

The upgraded platform has been a game changer for hayu, according to Andy Clarkson, Vice President of Technology Operations for the OTT service.

“The system provides increased editorial flexibility, allowing us to present a mix of editorially curated content and algorithmic recommendations (segmented and personalized) to users,” he said. “So far our data has shown significant increases to views of recommended shows.”

The Merapar team used another cloud service called AWS AppSync, to harness the benefits of using GraphQL APIs which enables increased performance and scalability. AWS AppSync is a fully managed service that makes it easy to develop GraphQL APIs by handling the heavy lifting and securely connecting to data sources like Amazon DynamoDB, AWS Lambda, and more.

“We commissioned a report to analyze hayu’s data and they mapped users to personas or segments,” said Birkinshaw. “We built a CMS that allows hayu to build lists of content and target them to users across the app portfolio. We started capturing a lot of live data from end users and built new personalization services that track their viewing history and generate new lists of suggested titles.”

The hayu service is also tied into a wider video CMS developed in-house at Comcast called MPX, which manages all video assets. Merapar has augmented that system with the management of metadata for enrichment, categorization, and list curation. This enables unlimited access to a wide variety of content titles. This was important to hayu because they have several extremely popular shows internationally as well as more specialized content for which they wanted to increase viewing.

Amazon Rekognition is used to find additional information on specific titles, like cast information, by analyzing content. This data is imported into Amazon Personalize to extend the recommendations, helping to surface shows with similar content.

Enabling its users to navigate the range of reality TV subgenres was a core goal for hayu. The business goal was to increase content consumption and gain a better understanding of their customers’ viewing habits and conversion rates.

“It’s all about flexibility and personalization,” said Clarkson. “The new platform gives us the ability to present users with a balance of editorial, segmented and personalized content, enhancing the overall customer experience. Increased viewing was the main aim of the project and we are pleased with the results and how quickly we achieved them.”