Part of:

How AI Is Personalizing Entertainment

Why Trust Techopedia

We've come a long way from the days of broadcast TV. Now we can watch what we want when we want, and artificial intelligence can help us pick out what to watch next!

From offering simple streaming content to highly personalized streaming content, streaming media technology has come a long way with a lot happening along the way. Offering streaming content was an achievement, but its progress was constrained by various factors such as computer hardware cost, limited computer capabilities, limited internet bandwidth and lack of compression technologies.

Then things changed as computer abilities improved, hardware and storage cost reduced, compression technologies were improved, internet bandwidth improved and that gave a boost to streaming content. Various events began to be broadcast via streaming and the results were encouraging. Businesses sensed a good investment opportunity and jumped onto the bandwagon. But something even bigger was to come – the mobile device revolution and artificial intelligence (AI). With easy access to high bandwidth and powerful mobile devices, customer expectations started to trend toward tailored content, and streaming content providers have been heavily using AI to offer people exactly what they want. (More and more people are giving up cable in favor of other services. Learn more in Cutting the Cord on Your Cable TV.)

How Did It All Start?

Believe it or not, the original idea to provide streaming content began way back in the 1920s, and that too, for a commercial audience. It came in the form of Muzak, a technology designed to stream audio to subscribers through electrical wires, rather than using a radio. It was not a great success, but the idea did not die. Rather, it would slowly consolidate.

Over the next several decades, computer capabilities improved, software and hardware costs reduced, internet bandwidth improved across the globe (albeit inconsistently), people had easier access to the internet and computers, and the idea of streaming content started moving again. Various developments charted the course. Both Microsoft and Apple created proprietary formats for streaming. Events were streamed live, which people watched on their computers. However, there was always this endeavor toward a single, unified, streaming format, and that's where Adobe Flash came in. Adobe Flash was used by many video hosting sites such as YouTube, which now defaults to HTML5. So, streaming content now had become almost seamless.

How Did Streaming Content Become Popular?

In 2017 Horowitz Research, a market survey firm, found that 70 percent of content viewers used streaming content and 40 percent of TV viewing was based on streaming. Millennials streamed 60 percent of the content they viewed. Clearly, offerings such as subscription video on demand (SVoD) and over the top (OTT) applications had become popular. This brought about the beginning of the decline of the DVD industry. In 2015, a New York Times report stated that Netflix’s DVD subscriber count had plummeted significantly, while their streaming service's subscriber count had increased. In March 2016, a study found that content consumers did not find significant qualitative differences between the DVD and streaming content. Clearly, the balance was shifting toward streaming. However, customers now needed features like fast forward, rewind and search. On top of that, advertising revenues had been soaring too. Clearly, content providers had incentives to invest more in streaming services. (For more on SVoD, check out How Far is Your Data From Your Analytics? An Overview of the SVoD Analytics Landscape.)

How Did Content Personalization Come About?

Interestingly, as streaming content became popular, content personalization had already been identified as the way forward. A content recommendation system was at the core of personalization. For example, Netflix, which was into the DVD rental business, had already been using personalization tactics that became more sophisticated over time. In the mid-2000s, Netflix offered cash incentives to people who would help improve the efficiency of its recommendation systems. Initially, the recommendation system helped viewers identify suitable DVDs. Over time, it would offer SVoD and OTT content based on individual preferences, browsing patterns, habits and other user inputs.


How Is AI Personalizing Content?

Identifying individual user preferences is enormously difficult. Think of, for example, Amazon Prime and Netflix with their enormous and varied subscriber base across continents and the personalization challenges. Additionally, subscriber behavior may suddenly change. You may also need to think about user experience, user-specific UI and many more such complex factors. AI and machine learning are capable of learning from subscriber behavior data over time and offering content recommendations accordingly. The system is known as a recommendation engine, and learns about subscriber behavior deeply on its own, like a human being’s learning evolves and adapts to dynamic subscriber expectations.

Netflix Content Personalization Case Studies

Perhaps no one has personalized entertainment content better than Netflix. Netflix allows its content to be driven solely by subscriber choices. Here is an overview of how Netflix goes about its job.

  • Custom subscriber experience – Netflix has more than 75 million subscribers in more than 90 countries and each subscriber is offered a unique experience or content offerings upon login.
  • Generalized recommendations – Since Netflix has huge and varied subscriber data, it may offer content recommendations based on criteria such as “this is enjoyed by people of your age group” or “your friend just watched this.”
  • Not offering an overwhelming content experience – Netflix wants to offer a recommendation its subscriber likes within 90 seconds, otherwise, the subscriber may move on. So, it does not drown its subscriber amidst a deluge of content recommendations and tries to offer just what you like. According to Chris Jaffe, Netflix’s vice president of product innovation, customers hate to be overwhelmed, so content personalization is the key.
  • Do not over-personalize – Though Netflix offers what subscribers tend to like, it occasionally throws in something not related to the preferences. For example, a subscriber with a demonstrated inclination toward horror flicks may be offered a comedy.
  • Let the subscriber choices govern – It is not the product or any other team that decides on content choices, but the customer. According to Jaffe, “We work on constantly making that experience better and better. It’s a unique approach. In some companies [that are] evolving the product, the product team might be the driver: The team comes up with the idea, design, builds, launches, and sees what happens. My team can’t make that decision. We come up with the ideas, but what drives product decisions is our customers and what customers do and how they use the product.”
  • New products must be approved by customers – All new products are run by the customers in tests, unbeknownst to them. Netflix runs “a couple hundred” product tests with 300,000 users each year. Only products with enough acceptance actually come to the shelf.

Impact of Content Personalization on Business

Netflix claims that it could save $1 billion by ways of decreasing churn. Monetary gains apart, Netflix could improve customer experience by optimizing data transfers and compressing codecs with a technology called Dynamic Optimizer. It has also optimized OTT delivery by preventing congested links and unstable connectivity by adopting Internet Protocol.


While there seems to be no doubt about AI’s potential, things don’t seem as good on the privacy front. The act of capturing customer data without explicit consent is a thorny issue, with many criticizing the practice. But that is a topic for another time. Meanwhile, customers can expect more interesting content on their watchlist thanks to AI.


Related Reading

Related Terms

Kaushik Pal
Technology writer
Kaushik Pal
Technology writer

Kaushik is a technical architect and software consultant with over 23 years of experience in software analysis, development, architecture, design, testing and training. He has an interest in new technologies and areas of innovation. He focuses on web architecture, web technologies, Java/J2EE, open source software, WebRTC, big data and semantic technologies. He has demonstrated expertise in requirements analysis, architectural design and implementation, technical use cases and software development. His experience has covered various industries such as insurance, banking, airlines, shipping, document management and product development, etc. He has worked on a wide range of technologies ranging from large scale (IBM…