Android Developer Story: CastBox improved engagement with Android Instant Apps & TensorFlow

Android Developer Story: CastBox improved engagement with Android Instant Apps & TensorFlow


Hi there, this is Xiaoyu Wang, co-founder and CEO of CastBox. Two years ago, when I made the CastBox prototype as a personal developer, I could not have predicted the podcast market would grow this rapidly. Google’s technology and platforms has given CastBox the chance to evolve and grow on a steady pace. Now, it is actively providing for over 7 million users the contents they love. The team also grew from a single person to a group of 40 in less than two years. Since launch, the quality of our product keeps improving, and the app’s review has been great. Not long after CastBox was launched, a blind user from Russia even sent us a thank-you letter, stating that millions of audio contents helped to bring her information across the world. Fascinatingly more, CastBox received Google Play Awards 2017 Nominee for Standout Startup, and as a result, the public got to know who we are. My name is Xiaocong He, co-founder and CTO of CastBox. My job is to optimize content value by utilizing technology. TensorFlow provides multiple algorithm model, and made our recommendation process more precise When user first log in CastBox, they can customize their initial recommended contents based on their interests. Then CastBox would combine listening data and user behavior, to best personalize the recommendation experience for individual user. Content to user, the precise two-way paring, drove CastBox to have a 70% Day 2 DAU Retention Rate. The average daily usage time reaches over one and a half hours. Also, with Android Instant App, people can use CastBox without it being installed on devices. This dramatically decreased the barrier for users to stay connected with content. I’m Shibin Li, co-founder and CMO of CastBox. The full realization of CastBox also requires engagement with content creators, so that we can establish a healthy content eco-system. We provide two monthly payment models based on Google Subscriptions: The first being channel-driven payment; Second being separate subscription based on functionality. The flexible subscription solutions not only provide directed services towards our heavy and normal users, but also give us insights to understand deeply about their needs. Google’s technology has provided a profoundly solid foundation for our success. Over 80% of our users come from the Android platform. And we are dedicated to bringing more functions to our users, provide the best catered content recommendation and the optimal listening experience.

19 thoughts on “Android Developer Story: CastBox improved engagement with Android Instant Apps & TensorFlow

  1. The sign in should auto detect if Facebook is or is not on the app and go to gmail like Amino app dose

Leave a Reply

Your email address will not be published. Required fields are marked *