Dynamic Paywall: Optimizing conversion with a machine learning model (Round 3)
Summary
In recent years publishers established paywalls and are now seeing promising results. The technology and models applied by all publishers are most of the time rule-based. Rule-based models however are too rigid to tap the full potential. With this project Saint Paul will overcome the limitations and establish a paywall 2.0: A dynamic paywall that automatically optimizes the settings based on machine learning algorithms in order to find each user’s individual tipping point to convert.
The solution
Today's e-commerce is highly personalized and automated. What is true for e-commerce isn’t true for publishers. Paywalls are still rigid and settings are configured manually. This does not allow for much personalization and means a lot of manual effort for the publishers. Saint Paul will set up a project team to build and optimize a dynamic paywall. A machine learning model shall help to determine the individual tipping point where a user accepts to take the registration and payment hurdle, both to achieve maximum conversions and reduce manual configuration work for the publisher.