Academy blog

Personalization is the New Frontier for Media Companies

The realm of behavioral analytics technology is a constantly evolving space. This is due to the innovative ways each company is learning more about users. From favorite genre of TV show to how long we normally watch for, the assistance of data collection offers uncomparable amounts of value to a streaming or satellite service.

Recently though, the depth at which a company looks for data has been expanded and progressive personalization engines have been employed by a number of companies to create a completely unique user experience, capable of keeping users watching through new and innovative ways to retain customer attention. Not to be confused with recommendations, personalization is a seemingly simple tactic that uses complex methods to analyze user preference and form a unique user experience capable of capturing the attention of users and harnessing more data using machine learning.

The difference between recommendation and personalization


Recommending content is a very commonly used technique to not only keep viewers watching, but also to learn about users’ likes and dislikes. This technique involves a media provider offering more content once a customer has finished a video. Once the user completes their view, it’s expected that the service will want to keep hold of the user and maximize watch-time. Now, offering a few random or popular shows is doing some good but it’s really stabbing in the dark. What a recommendation engine will do is remember the type of content people who watched this video normally chose next – this way your chances of catching their attention grows exponentially.

There are different recommendation algorithms defined by one or a combination of elements, such as:

  • Active contribution from users: thumbs up or down, or starts.
  • Auto-rating based on video completion rate.
  • What similar users watched (what configures a similar user type may change: consumption patterns, demographics etc).
  • What you recently watched (based on content description).

For example, the most successful and well-known use of this will be Amazon. Their recommendations bar or “What other people bought after viewing this product” slideshow accounts for 35% of their revenue and continues to be a shining example of how you can use big data to learn about user behavior. Let’s say you’re looking for a new Television. When you’re viewing a product, the recommendations bar will use the data from other users to show the most common timeline of behavior that led to a sale and attempt to replicate this. As an extra boost to their revenue, once the purchase has been made, Amazon then recommend related products that others had bought after the original purchase. It’s a clever and simple method of using data analytics to transform your content offering.

Some companies might want to use an industry leading content recommendation platform such as Spideo in order to guarantee they’re in line with industry standards, but it’s entirely up to you as to how you prefer to utilize the tool.


While recommending content is the use of data to discover the most popular content and push it towards users, personalization is done on a user-by-user basis, using A/B testing and behavioral trends to learn not only what a user likes more, but what they liked about it.  The more a company knows about someone, the more effective it becomes. This explains the appearance of personalization as a marketing tool. This method uses highly-intelligent learning algorithms designed to not only read use behavior and feed it back to you but use machine learning to predict to a level that recommendations just can’t compare.

Sometimes referred to as one-to-one marketing, the full potential of this method is not yet known, however many businesses are using this open space to experiment and create powerful results. Personalization is getting better and better, particularly among big, leading video players, and existing holistic, BI solutions now provide deeper insights across the entire video platform to correlate UI/UX, recommendation, advertising, and many more, with customer behavior to get closer to a fully engaging user experience. Yet, this is made harder by regular changes to privacy laws which are constantly being revised. Check out the NPAW guide to GDPR here.

Personalization will put you a league above the competition using NPAW

As mentioned above, the full potential of personalization hasn’t been realized just yet but there are definitely two very well-known media services who are leading the way…


The Powerhouses


The home screen of Netflix - an example of personalization

Now with over 125 million subscribers, there’s an extremely high chance you’ve used Netflix before or at least seen their unique home screen. The first impression from this is just how the website capitalizes on real estate to create an extremely visual experience. But Netflix isn’t just doing this to look pretty…

While someone is interacting with your website, it’s in your best interest to learn as much as you can from each and every move they make. Netflix has invested a lot of time and money into categorizing their promotional imagery and linking different representations of shows and movies to someone’s taste. For example, here’s how Netflix describe it on their blog.

“Let us consider trying to personalize the image we use to depict the movie Good Will Hunting. Here we might personalize this decision based on how much a member prefers different genres and themes. Someone who has watched many romantic movies may be interested in Good Will Hunting if we show the artwork containing Matt Damon and Minnie Driver, whereas, a member who has watched many comedies might be drawn to the movie if we use the artwork containing Robin Williams, a well-known comedian.”

This helps to illustrate that difference between recommendation and personalization we discussed earlier. As Netflix’s engine learns each customers habits, it’s understanding how to catch their attention, how to persuade them to click and how to keep them on their page.

Once they have designed the marketing imagery to match the various categories they put users into, Netflix then uses their own parameters to measure how to sort a user and do this through a process of A/B testing and popular data much like recommendation. To understand their users it takes time, patience and a big budget to make sure they know every user on a one-to-one basis.


Almost 5 billion videos are viewed every day on Youtube – that is 35 times as much as Netflix. Youtube is an industry juggernaut and after creating a space completely commanded by user-generated content, they made their move towards being a Transactional Video on Demand (TVOD) service with Youtube Red (Now called Youtube Premium).

The industry leaders have always sold themselves as a platform geared towards promoting creators, however, as the platform grew into an extremely lucrative space for advertising and paid content, Youtube has decided to tweak their user pages using personalization.

Normally published in a chronological order, Youtube has now made an effort to expand beyond their recommendations homepage and use A/B testing to learn more about how users interact with the channels they’re subscribed too. This is in response to complaints about the recommendations page pushing content that doesn’t apply to them.

This issue with Youtube is that with 300 hours of video uploaded every minute, it’s not as simple a task as Netflix has. Youtube is still in the very early stages of experimenting with personalization, but with such a huge and successful team of analytics experts behind them, you can be certain that this move has been pre-meditated and rolled out with a good idea of what results it will bring in.

Why use A/B testing?

  • Better experience for the customer
  • Focused recommendation for better conversion rates
  • Improved general content, better website ranking
  • Greater return on investment (ROI)
  • Building customer loyalty
  • Focused Email campaigns
  • Improved leads and ranking
  • Reduce risk of user churn

At the core of both recommending and personalization is user data. To gain full visibility of how your content performs and to learn more about your users, click here to discover YOUBORA, our decision-making solution.

Research & Editorial Team on June 06th 2018

Most Recent Posts

Introducing the YOUBORA Mobile App

As the streaming market matures and user expectations surge, the ability to immediately detect issues and share up-to-date reports across an organization is becoming an essential part of the day-to-day work lives of video industry professionals. In response, NPAW has launched the YOUBORA mobile app,... 

10 Jun

How to Weigh the Impact of Offline Downloads on your Streaming Service

As streaming becomes an ever-more established behavioral trend globally, online video providers are faced with a double-edged sword. Demand for online video is higher than ever, but so too is competition for space in consumers’ viewing habits. User expectations are on the rise and, as it is easier... 

3 Jun

How COVID-19 is Changing Online Streaming Behavior Globally

The outbreak of the COVID-19 virus is drastically changing user behavioral patterns worldwide. As the virus spreads, streaming video providers must contend with sudden changes in variables including rising economic pressures, altered routines, moods, and priorities. We have gathered together a series... 

27 Mar