We here at The Filter are involved in many pitches and we have monthly insight sessions with all our clients. If there is one key aspect that we have found works either to persuade a potential client that recommendations really move the dial or to show an existing client that they need to increase, say, recency in their models, it is data points. We have data at the heart of our business, not just because it is the basis of all the data science we do, but because we don’t exist unless we can show how personalisation can move the dial for our DTV and OTT partners. Below are some real-world data points that we find amazing about the subject we love – personalisation.
“For You” drives 3x more conversion than “Most Popular” – for one of our clients, we recently tested a variation for one of their top content rails. For 90% of their base, we carried on showing our “Recommend for You” which was tailored to that user’s viewing history and time of day. For the other 10% we showed the ‘Most Popular’ rail which was strong but not personalised. We were expecting a slight uplift as their ‘most popular’ had some great titles in there. But a 3x conversion uplift was even beyond my optimistic projections and demonstrates the effectiveness of personalisation.
“More Like This” (MLT) added 15% extra PPV revenues – MLT is based on content-to-content recommendations. To some degree consumers are more willing to accept it as they can believe that a machine could ‘know’ content but find it harder to accept that a machine can ‘know’ them. However, MLT tends to be buried at the end of a content discovery journey just when a user is about to press play on some item of content. So how much does MLT at this point actually move the dial? We were pleasantly surprised when we worked out that for one of our movie and TV PPV services, this ‘buried’ MLT was actually adding an extra 15% to the total revenue of the client’s service. We subsequently tried to convince that client to test “Because you Watched” on their home page which is basically an MLT based on that person’s recent viewing history. It really worked.
32% bounce rate is best in class – I am often amazed how many potential clients don’t measure bounce rate. (We here at The Filter define bounce rate as percentage of users who come to your platform and do not start watching anything within 1 hour). For any e-commerce site, bounce rate would be one of the core metrics being tracked. It is such a critical measure for DTV and OTT platforms to get a feel for how well their overall UX is performing. For our clients we see a large variation in this metric – the average is just under 50%, with the worst being up near 80% and the best being down at 32%. The Filter offer bounce rate tracking as a core part of our managed service. We also track ‘Visits in last 30 days’ and this can vary from between 2 to 6 on average across the entire base. This is also a key metric and one which personalised customer marketing can really help to improve.
25% total plays from recommendations is current average but rising fast – there are two extreme schools of thought out there; one says that you should use editorial teams to drive all your UX content rails as they really get content and what your users could start liking and the other says that data science should power the entire UX as it can get to know individuals and optimise for their tastes. Clearly the answer lies in-between the two. We typically find a quarter of all views come directly from content rails that are powered by recommendations. The other three areas that typically drive significant traffic are:
- Hero banner – editorial teams often power this prime real estate
- Continue watching – whilst you could argue this is not editorial, we class it as not powered by recommendation
- Search – whilst we do tailor our search to make it somewhat personalised based on that user’s history, search is not a direct recommendation
Here at The Filter, we strongly believe that combining the power of editorial teams and data science is the sweet spot. We even offer a HERO tool that allows editorial teams to select a bucket of top titles and then the power of our AI is used across just those titles. One of the biggest wins here is simply removing titles from a hero banner that have already been watched by that individual.
High churn is linked to watching 3 or less content titles per month – when we have had the opportunity to look at churn for our clients, we often see that the biggest driver of retention is content engagement. This might sound obvious as when people don’t watch anything on your service, they clearly don’t value it! However, it is more nuanced than this. Above 3 titles per month, it does not matter how many shows that person watches; their churn profile is low and driven by other variables like streaming quality and customer service issues. However, below 3 titles per month and their churn risk becomes significantly higher. All clients need a plan to move these low usage viewers into engaged users. The Filter’s personalised marketing products are one powerful option here.
Of course, every client has different customers, and every client has varying business goals. The Filter offers a managed service model for recommendations/ personalisation which means that we adapt our data science to your particular needs. We provide all our analytic dashboards out the box for all our clients as this is the quickest way to provide common understanding and focus efforts. We also continually run a/b experiments for our clients to see if we can nudge the metrics in the right direction. We hope that the above will give you some hints in understanding where the opportunities reside for your business.
For more information please email firstname.lastname@example.org
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 means more loyalty, lower churn and more revenue.
The Filter prides itself on working closely with clients both during set-up and afterwards to make sure that the client’s customers are always getting the optimal recommendations. We are an agency not a SaaS provider. We have a genuinely deep TV understanding which means our many 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. We also integrate effortlessly into your data ecosystem allowing you to benefit from the power of personalisation easily and quickly.