ML and the Use of Smart Data to Attract, Acquire & Retain Subscribers in OTT is one of the most relevant topics nowadays. The media OTT world is getting increasingly crowded. Disney’s video-on-demand streaming service launch in 2020 attracted 10 million subscribers in its first 24 hours in just the US, Canada and Netherlands. Subscription video-on-demand platforms such as Netflix, Hulu, and Amazon Prime Video have accumulated millions of monthly active users and provide access to a variety of movies, TV shows, and documentaries at the click of a button.
There’s no doubt media consumption has increased substantially in digital formats, with the lockdown accelerating this trend.
- According to Statista, the average time that Americans spend on subscription over-the-top (OTT) video will increase year by year to a projected 70 minutes a day in 2022.
- The average viewing frequency of consumers is around 12.5 times per week in India according to KPMG.
- A Juniper Research report estimates that in 2020 there was an average of 4 SVOD subscriptions per household in the US, but with growth slowing significantly from 2021.
- The report also estimates that over 70% of streamed video sessions in the next five years will occur on smartphones.
And the forecasts continue to be really optimistic for the coming years. Some media analysts predict that there will be nearly two billion active subscriptions to on-demand video services in 2025, representing an increase of 65 percent over the end of 2020. The primary engine for this growth will be traditional broadcasters, who are increasingly turning to streaming services as a way to extend their reach and compete with online video giants such as Netflix and Amazon Prime Video. The closure of cinemas, theaters and even some sports events also obliged some businesses to digitize their product and broadcast this content live or on demand on these platforms.
This kind of growth brings with it a lot of opportunities and challenges for video service providers and the unprecedented uptake and consumption of OTT video services and other digital media and entertainment are leading providers to utilize smart data enhancements such as artificial intelligence (AI) and machine learning to fully understand their customer base.
While the big players are definitely responding with a wide variety of content and distinct UI/UX experiences, they are facing stiff competition from several regional OTT players. In some cases the market share of these regional players is even surpassing that of the big players like Netflix and Amazon Prime.
A key differentiator in the success of these players is their understanding of their audience and the ability to deliver relevant content. User behaviour analysis and AI/ML-based personalized recommendations play a significant role here.
The success of any OTT platform depends on content, discovery and user experience. Personalization in OTT platforms works on all these factors and simultaneously helps OTT brands address marketing challenges related to:
- Reducing the costs of customer acquisition and increasing CLTV
- Individualized recommendations of relevant content
- Increasing average user content consumption
- Trial- to paid-user conversion rates
- Increasing engagement and retention
- Reducing data processing costs and efforts
To address these challenges you need to have a deep sense of your customers’ journey, You need to know exactly WHAT is happening in your video service, WHY it is happening and WHAT’S NEXT in order to launch hyper-segmented campaigns and make an impact only on the specific group of users you want to address at a specific time with a personalized offer.
With that purpose in mind, we will explain how to use data smartly during the acquisition, engagement, and retention phases:
In this fierce market, understanding our customers’ journey is key to attracting new ones, but retaining customers for as long as possible is even more crucial. It is common knowledge that it is much more expensive and requires a much greater effort to attract new customers than to build loyalty among the ones your already have; and loyalty is built only by engaging them with the video service by providing them exactly what they want, for a truly great viewing experience.This cannot be accomplished without the smart use of data to measure the insights that have a real impact on your entertainment business ROI and improve your customers’ experience.
Acquisition is one of the toughest challenges for any company, but the old rule of thumb that it costs 5 times more to get a new customer than it does to keep an existing customer is not true anymore. In the past we didn’t have predictive analytics to accurately guess future behavior that we now have thanks to advances in ML and AI. The selling process has greatly evolved now, and the buying process is focused on fulfilling customer expectations. So personalization is king, because of its ability to exponentially increase consumer spending. The focus in the acquisition process, in line with the Blake Morgan philosophy from the well-known customer experience futurist, should be on connecting with customers and delivering value – now and in the future. The most successful companies find a balance between the two costs. When considering how much to spend on capturing and retaining customers, it’s important to consider customer lifetime value (CLV) and projected CLVs. Knowing how much your customer is worth can help you make smarter, more accurate investments rather than spending a lot of money all around to prevent churn.
Acquisition in the video industry has become a bit easier during and after the lockdown, with a huge rise in the number of subscriptions that is expected to continue in the coming years all over the world. Most households have more than one subscription but the drivers for customers to subscribe to an entertainment platform are very diverse, according to Parks Associates in their report on the main influencers for subscriptions in OTT platforms.
Knowing exactly how your potential subscribers get to your video service is key to determining the most profitable channels to invest in for new customer acquisition.
Knowing how many new users you have, which service or package they subscribed to, and when and how they subscribed is fundamental to forecast your future revenue and plan the actions you will take ahead of certain situations.
Knowing the demographics and being able to predict future behaviour will also help you launch hyper-segmented campaigns and tailor your impact to specific groups of target users.
- UAP. User Attribution Performance is another fundamental KPI to know how your acquisition channels are performing and to be able to personalize your acquisition campaigns. With this data you can know your service conversion rates for trial users, paid users by conversion channel and paid users lifetime value per acquisition channel. This will let you predict your future conversion rates and launch specific campaigns to improve these numbers.
It doesn’t matter if we are talking about gaming companies, SVOD, AVOD, transactional-based SERVICES or a hybrid. The priority of any company after attracting a new customer is to secure its loyalty and keep it engaged to the service as long as possible, so satisfaction here plays an important role. To retain and keep your users engaged you need to make their experience on your platform as smooth as possible, make content discovery as easy as you possibly can and recommend your users exactly what they want to consume at any specific time.If they find value in your service, they’ll stay.
In order to offer the best customer experience, you need to fully understand your users needs and wants, track their behaviour on your platform, be able to predict their next actions and offer them an individualized experience. Machine Learning and Artificial Intelligence can be really helpful here, to understand your users’ behaviour, segment them, and predict their future actions.
Knowing the demographics and tastes will help you profile each user, understand their journey with your service and keep them more engaged by offering exactly what they want.
The ability to know how your users are grouped based on their engagement level, or to create clusters according to the type of content they like, the genres they usually watch or by taste communities, is also really helpful and lets you implement very specific campaigns, which allows you optimize your marketing investment by targeting only a specific cluster of users. Imagine you have a new drama series and want to approach only the drama fans that have been inactive for the past month in an attempt to re-engage them.You could just select that cluster and impact those users through, say, an emailing campaign or push notifications.
If you succeed with content recommendation you can be sure that engagement levels will increase dramatically and consequently so too will the time users spend on your platform consuming content,and this in turn will lead to a greater loyalty level with your service and a higher CLTV.
Knowing which content performs the best – or worst- and which content providers are the most successful can also help you optimize your investment in content.
This deep understanding of your video service user’s data will assist you in providing them a personalized experience. ML and AI let you offer contextual recommendations for your users tailored to their consumption habits and personalized recommendations adapted to the different user consumption scenarios to increase user satisfaction and engagement.
Jump predictions help you boost acquisitions by using predictions to identify the most likely candidates to convert to paying customers and lets you leverage AI models to predict the trial users that are most likely to convert, thus defining a solid conversion strategy with immediate, concrete actions.
Engaged customers will turn into higher CLTV and this will have a direct impact on a company’s ROI, so customer retention is fundamental for any company and again, knowing exactly what your users want and how they behave in your video service is crucial to predicting when they are most likely to leave your service.
There are many drivers for churn but the good news is that with Jump predictions you can measure the KPIs that put your users at risk and impact them before it’s too late!
You can increase retention by understanding why your users would leave, and before they leave by leveraging specific AI models to predict churn and understand why customers might be at risk, leading you to a retention strategy with defined concrete actions.
Churn is predictable and the precise knowledge of our users lets us accurately predict when they are more likely to leave, so that we can reverse the situation, get them to stay and increase CLTV.
In conclusion we can say that service discovery, acquisition performance, audience engagement, service monetization, churn analysis and many other behavioral metrics, taken altogether (as it wouldn’t make sense to base business decisions on any one metric) will help your OTT business jump to the next level by using data smartly to understand your user base, predict user behaviour, enhance the effectiveness of your marketing activities and as a result optimize customer acquisition, retention, and engagement.
Download our Whitepaper on “ML and the Use of Smart Data to Attract, Acquire & Retain Subscribers in OTT” to learn more.