How Video Streaming Services Can Reactivate Lapsed Users With Deep Learning
With an annual growth rate of 20.4 percent from 2020 to 2027, the global video streaming market will be worth US$184.3 billion by 2027. Netflix is still the leader in this sector with 167 million paying streaming subscribers worldwide, but its market share has fallen significantly from 91 percent in 2007 to just 19 percent last year due to increasing competitions from the likes of Hulu, Amazon Prime and Disney Plus.
Regional streaming services are also seeing a further leap in viewers and subscribers. China’s streaming service iQiyi passed 100 million subscribers in 2019, while TikTok – a Chinese video sharing and live streaming platform – has become a global phenomenon with over 1.5 billion total downloads as of last year. Its audiences open the app eight times a day on average.
There is no doubt that the sector will continue to grow. Apart from increasing high-speed internet access across the developing world along with social media usage, advanced technologies such as blockchain and artificial intelligence (AI) are the key drivers behind this growth.
However, with more than 400 subscription-based video-on-demand services battling for the attention of a global audience of 1.1 billion, publishers and video streaming services need all the help they can get to retain their precious viewers.
Why Are People Leaving Your Platform?
User churn and retention has always been a major hurdle for video streaming services. A recent study shows that 67 percent of app users become inactive in the first two weeks.
At a basic level, viewers don’t return because something about their experience wasn’t good enough. This could either be technical or content related. Other reasons could also be rising prices, changes to their packages, or a must-see series or sports season that has ended.
Bringing back those inactive users is critical to your business success. A 5 percent increase in customer retention can lift profits by 25-95 percent, according to Bain & Company. However, that doesn’t mean you should spend money and time trying to win back every lapsed user. Just reach out to those valuable ones, something that AI is well placed to assist with.
How Deep Learning Can Help
The most valuable inactive users are those who are most likely to be brought back. So, how can you find them? Traditional behavioral targeting is difficult as it requires human intuition to draw extensive insight from data points that are easy to misinterpret. However, deep learning, an advanced AI technique, can conduct multi-dimensional analysis of historical user data on behavior, frequency, time, devices they used, what they watched, etc. to identify and segment inactive viewers. It can then predict conversion rates (in this case, the likelihood to be reactivated) of over one million segments to find out the most valuable ones.
To ensure your targeting with the highest precision, you also need to know inactive users’ current interests. By combining historical data and third-party data, AI can discover users’ external interests outside your platform. For instance, Emma used to watch a drama series on your platform, but she has been inactive ever since the show ended. By analyzing her browsing behavior on external websites, you find out her recent interest in comedy as she has been reading reviews about comedy series and stand-ups. With this information, you can then personalize your campaign with the most relevant recommendations and creative to win her back.
A leading video-sharing and live streaming platform used Appier’s AI solution to win back users who had been inactive for over seven days in Taiwan. It managed to reduce cost-per-click by 22 percent and cost-per-acquisition by 7 percent compared to the platform’s own KPIs. By using dynamic creatives, it also saw the conversion rate of its native ad jumped 30 percent. Overall, 10,000 lapsed viewers returned to the platform and each watched three videos on average within one week.
By serving your inactive viewers in ways that properly align with their behavior and interests, you are far more likely to see them return, stay engaged, and become loyal customers.
* Struggling to maintain a low churn rate for your OTT video subscriber services or video streaming platforms? Not sure how you can bring those lapsed viewers back? We are here to help! Our AI-driven solution powered by deep learning enables you to identify who are most likely to come back and what kind of content will win their hearts. Get in touch with us today for an exclusive discussion.
WE ARE HERE TO HELP
YOU MIGHT ALSO LIKE
Artificial intelligence (AI) is heralding a revolution in how we interact with technology. Its capabilities have changed how we work, travel, play and live. But this is just the beginning. The next step is explainable AI (XAI), a form of AI whose actions are more easily understood by humans. So how does it work? Why do we need it? How will it forever change the way industries – especially in marketing – function? The Mystery of the Black Box: The Problem With Current AI No one would deny that artificial intelligence produces amazing results. Computers that can not only process vast amounts of data in seconds, but also learn, decide and act on their own have turned many industries on their heads – according to PricewaterhouseCoopers, the market worth of AI is around US$15 trillion. However, in its current form, AI does have one major weakness: explanation. Namely, it can’t explain its decisions and actions to humans. This is sometimes referred to as the “black box” in machine learning – for example, the calculations and decisions are carried out behind the scenes with no rationale given as to why the AI arrived at that decision. Why is this a problem?
Rich Wilson started his career writing software for jet engines. Since 2005, he has been focusing on the commercialisation of software startups. He has a history of successfully bringing US tech firms into Europe and has since founded a number of businesses including Volcube, a financial services tech firm, a small hedge fund which uses artificial intelligence (AI) to place trades, and most recently Deviance, an audience diagnostics firm applying AI to language analysis in order to help firms understand how their audience feels. Rich is based in London and is a noted speaker on the application of linguistics in politics, gender, TV and music, as well as tech startups more generally. We recently spoke to Rich to discuss the current adoption of AI and data, the impact of technology on some of the key industries, and how to strike a balance between data privacy and customization. You have founded and helped build a number of AI and data businesses, including in the marketing space. Tell us a bit about your experience and what you’re working on now. I have a background in software development, and during the dotcom boom I built some of the first online stores for
To solve the problem of disconnected data in a world where we are dealing with millions of touchpoints and interactions, 51 percent of marketers say they currently use a customer data platform (CDP) and 35 percent say they plan to use one. So, what exactly is a CDP, what benefits does it offer, and how does it compare to other customer data systems? What Is a Customer Data Platform? A customer data platform is a type of software that collects and organizes data from a wide range of touchpoints in real time to create individual, unified customer profiles. CDPs can also communicate with other delivery and engagement systems such as marketing automation software, websites, and social media management tools. This means it is possible to plan and execute campaigns directly from them. Some customer data platforms also have additional functions built in, such as business intelligence, analytics, reporting, and tracking. Their ultimate purpose is to help businesses and marketers make better customer-focused decisions. What Type of Customer Data Does a CDP Work With? A customer data platform can collect and work with multiple types of first-, second-, and third-party customer data. This includes data from your CRM activities, web forms,