Freadom is a productive engagement app for children. There are multiple media types – books, audio, video, games, etc. For example, when a child launches the Freadom app, she chooses her book (or other item) from a handful of cover images displayed on the screen. An alternative behaviour would be to use the search tab – but we find this atypical.
These ‘displayed’ stories represent only a few possibilities from a vast pool of stories. It is obvious that we can only display a few titles since the area available on the screen is limited.
The problem of picking the right title to display as per the user is a common problem across similar media apps like Netflix and Spotify.
The recommendation engine governs story discovery. A child chooses what to consume based on what she is shown by this engine. Therefore, her progress in the app, her learning outcomes, and her satisfaction depend on how well the recommendation engine works.
The Freadom app was launched in 2016. The recommendation system has gone through multiple iterations since. However, 23% of sessions on the Freadom app end up with a child skipping all presented stories. Additionally, we know that a typical user starts four stories before completing one (on average). One wonders what this data looks like for Netflix. Nevertheless, there is visible room for improvement.
We have partnered with Golub Capital Social Impact lab at Stanford University which is lead by Dr. Susan Athey. Users tend to have differentiated tastes, thus providing bespoke recommendations can substantially boost their engagement and improve the learning pace. Perhaps one child likes animal stories, while another one prefers stories about sports; A third child may like stories with pictures, while a fourth one prefers moral stories.
Recommendation Engines -> App Success
Organising and displaying content as per individual tastes is the holy grail of online platforms – in ecommerce, media, and almost every other sector. Many clever recommendation engines are actually a combination of various algorithms. The better the recommendations are, the more the engagement on the app. We have seen this in Tiktok recently.
The highly personalized modern-day Facebook and YouTube feeds are examples of these sophisticated recommendation systems. Personalized recommendation engines leverage methods from Computer Science and Statistics to construct tailor-made recommendations based on users’ past behavior. These models attempts to pin down important title characteristics which drive choice and figure out users’ tastes for these characteristics based on user’ past interactions.
You can download this short executive summary to read more about the recommendation engine.Rec-engine_executive_summary
A productive screen time app for ages 3 to 12, that focuses on improving English Language skills.
Online English classes for ages 5 to 12. Proven methods for children to improve academic performance and confidence.