Mainstream media is highly competitive. With so many news and entertainment options, getting a consumer’s attention—let alone his or her subscription fees—can be an uphill climb. Pythian helped a global media company become more targeted and efficient at attracting subscribers by building a custom machine learning (ML) model and content recommendation engine.
Together, these powerful tools leverage the company’s robust data to predict a website visitor’s propensity to subscribe. With these insights, the publisher can now tailor future online content, marketing campaigns and special offers, all to convert more individuals from casual browsers to paid customers.
The company had a vast repository of audience data but struggled with converting viewers from casual browsers to paid customers. They wanted to implement practices that were robust and flexible enough to expand across all their digital media titles. While their internal Business Intelligence (BI) team knew an ML model was the answer, they looked to Pythian to provide expertise and guidance on the model’s design and implementation.
Pythian collaborated with the publisher to outline their strategy, then began data analysis and feature engineering, leveraging tools including BigQuery, TensorFlow and Jupyter Notebooks based on Pythian’s familiarity with Google cloud services. Pythian also employed Google services to build a custom content recommendation engine to give business stakeholders an on-demand way to test variables that would impact subscription rates. The business is now well-equipped to optimize their marketing, operate more efficiently, and ultimately increase subscription revenue, using Pythian’s custom ML mod.
