Why Live TV Is Important for Recommendations

Some of the most enjoyable meetings we have with clients are the kick-offs – we usually get to understand the vision a client has for where they want their customer experience to go and the role that personalisation and recommendations can play. We also get to add in our thoughts about what we see working and what will not. Great collaborative sessions.  

There is one clear trend emerging in these kick-off meetings which we here at The Filter think is great for customers – clients are increasingly asking us to include Live TV as part of any recommendations experience. 

Lots of people assume that personalisation and recommendations are only for on-demand content. However, for many clients, live TV is still >50% of all viewing. So surely you should include it in any ‘Recommended For You’ or ‘Most Popular’ algorithm? If you have sports and news content, this will always be best live.  

But how do Live TV recommendations work in your CX? Recommending the best shows for a customer to watch from the entirety of next weeks’ live viewing is clearly not a great experience. So you have to link any live recommendations to what is on now. In practice, many of our client use countdowns for shows that are about to air i.e. in the next 5 minutes. We feel this countdown experience adds to the excitement. Many OTT players now have ‘start over’ functionality and so you can keep any live recommendation on-screen for the entirety of that show. Some PVR based clients link live recommendations to ‘set a recording’ and you can also ‘set an alert’.  

On a UI there are multiple ways you can manifest these live recommendations: 

    • As part of a ‘Recommended For You’ rail – if a customer loves tennis and The French Open semi-final is in 4 minutes, it must be part of the ‘Recommended For You’ rail.
    • Hero banner – we offer a tool whereby editorial teams can combine their selection skills with a data science overlay. The hero space looks wrong when it does not include live and so the editors should pick out the best live shows of the week and we display them in the hero at the right time; only for the right customers of course. 
    • ‘Live for you’ rail – this takes all the shows that are live now or have just started and orders them for that particular individual or household at that time of day. Basically this is ordering the EPG for you at that moment in time. If you always watch the Line of Duty at 9pm from channel Y, that would be first in this rail at that time. Very cool feature. 
    • ‘Most Popular’ – clearly live shows should be in this rail but how do you get the popularity data for something when it has not yet aired? We tend to find that using last week’s popularity data for a show works well. Trending is even harder still as it is based on rate of growth of views over a short time and is designed to pick out the new and upcoming shows not the established, big hitters. You generally need a third party data source to include live in trending, like twitter mentions. 

Content metadata is also often mentioned when people ask about live. In reality, it should be treated exactly the same as metadata for on-demand. There will be gaps and errors and it will need structuring in exactly the same way as for on-demand assets. Tags and franchises will also need adding and in fact the secret sauce to get very strong recommendations is using the same set of tags and franchises across all your content assets irrespective of them being live or on-demand.  

Channel brands are also an important addition to this metadata. People use channel brands as navigational signposts and even more so for live. So any recommendations should use them as part of the weightings. If you generally favour CBS more then you should be recommended more live shows from them.  

For any service that also contains sports and news, live personalisation is significantly more important. These live events can often be relegated to lower down the UI or customers are forced to use an EPG and scroll to them. Add in live to the hero and ‘Recommended For You’ rails at the top of the home page and this problem is somewhat solved.  

There is no reason why every rail in your UX cannot contain live recommendations for that person. If a significant proportion of your viewing is still linear channels, then you would be remiss for not including it. It is almost essential for services with sports and news on them. We here at The Filter feel this is a vital part of recommendations that is often overlooked in the rush to assuming everything will be on-demand one day soon. We also happen to offer live content as part of our recommendations.  


About The Filter 

The Filter offers personalisation of key customer touch points based on proven data science. We put the right content in front of a customer at the right time. More views mean more loyalty, lower churn, and more revenue. 

The Filter prides itself on working closely with clients both during set-up and roll-out to make sure that the client’s customers are always getting optimal recommendations. We are an agency not a SaaS provider. We have a genuinely deep TV understanding which means our data science solutions work exceptionally well for your viewers. We work with you to tailor our AI to your customer experience, your audiences, and your content. All integrated effortlessly into your data ecosystem allowing you to benefit from the power of personalisation easily and quickly, developing your platform at speed.   

For more information, please email damien.read@thefilter.com