New Research: Mobiquity’s Eduardo Barbaro To Predict User Engagement

Mobiquity’s Senior Data Scientist Dr. Eduardo Barbaro has received a full grant for his proposed research on user engagement with Dr. Ivano Malavolta of the Vrije Universiteit (VU) in Amsterdam. The grant has been issued by the Network Institute, an organization that brings together Amsterdam-based academic institutions and parties interested in their research findings and knowledge, and will further establish Mobiquity as a serious player among digital engagement providers.

Dr. Barbarbo’s research aims to formalize, implement, and evaluate a numerical framework for modeling user engagement in mobile applications to predict engagement behaviour. Why does this matter? Because over 95% of smartphone users stop using an app by the end of the third month after downloading it. This level of disengagement is just too high, especially when considering the growing impact that digital solutions have on a brand’s bottom line.

Focus on retention

While many app developers use the influx of new users as a measurement of success, Dr. Barbaro says that the retention of existing users is much more interesting. “I compare that first way of thinking to buying a cat and seeing yourself as a successful cat owner. But you didn’t buy any food, so the cat dies, and you are not a very good cat owner at all. Mobile applications must keep users active over a sufficiently long period of time to be considered successful. Predicting when users are likely to engage and disengage is very valuable knowledge, and the research will seek to provide such insight.”

Dr. Barbaro is excited to begin the research and shed light on this important topic. “Besides the interesting research, this partnership strengthens our connections with the VU and establishes Mobiquity as a serious player in the field of user engagement.” The final outcome of the collaboration will be published in an international peer-reviewed scientific journal, and learnings will inform our future client engagements.

Solving the challenge: User engagement in mobile applications

While user engagement in many contexts has been researched before, including massive open online courses and online health communities, understanding it in a mobile context remains a challenge in itself due to the fast-paced nature of mobile combined with the short lifespan of most mobile apps. Dr. Malavolta will work together with Dr. Barbaro and his team to further explore an algorithm developed by Mobiquity in 2016 for characterizing user engagement, specifically in mobile applications. This so-called clustering algorithm was designed to classify and group users based on their past behaviour and was then coupled to two different numerical models to perform the actual user engagement predictions. Mobiquity Data Scientists Esther Weusthof and Jeroen van den Hoven are also part of this research.