Understanding the science of habits gives media companies a better understanding of how customers engage with a brand, Patrick Tornabene, Newsday’s vice president of audience development and analytics, said Thursday in the INMA Webinar “Newsday’s Predictive Data Analytics Driving Subscriber Retention.” 

Habits are a craving, Tornabene said. A craving can take four days, or 100 hours to break. Any shift in a habit, such as a customer’s payment schedule, can be an indicator of potential churn.

“It’s really focusing on how the customer engages with the product,” he said. 

With predictive data analytics, application of models on aggregate records gives a holistic view of the customer’s relationship with the brand. Despite the knowledge driving predictive data analytics, customer actions must be interpreted from a customer’s point-of-view.

“It’s a careful balance between art and science,” Tornabene said. 

When a customer calls repeatedly to complain, one may assume they have a higher churn score. But, Tornabene said, any engagement is positive. A customer calling to complain cares more about the brand relationship than someone with less contact with the company.

This understanding of customer engagement habits combined with data analytics can help a company identify and slow potential churn. 

In four case studies, Tornabene demonstrated applications of predictive analytics on churn reduction and persuasion modeling. 

1. Dynamic messaging for engagement deviations: Changing payment habits can be indicative of change, Tornabene said. In this case study, Newsday saw a reduction in formers, customers who did not restart their subscriptions after lapsing drop 13% three to four months after launching dynamic messaging.
2. “Surprise and Delight” retention marketing: After identifying customers with a high churn score, Tornabene and his team sent different types of gifts and tracked the response. They reduced the former rate by 40%, and learned that a greeting card was the most effective item in both cost and response.
3. Optimising “Surprise and Delight” retention marketing: Building on the knowledge of the previous case study, Newsday worked to target what they determined was a “swing group,” those more likely to be influenced by this communication.
4. Determining niche product opt-ins: In this case study, the team recognised a demand for puzzle books. By offering the product as a free opt-in — then later increasing the price rather than pushing a paid product from the beginning — there were less complaints about the price increase. 

With a partner, news organisations or any other subscription-based company can get off the ground with almost no capabilities in-house, Tornabene said. 

Ned Kaufman, manager of data analytics at Newsday and a panelist on the webinar, said in-house skills begin with a basic understanding of data and some familiarity with databases to connect the data. 

Adding to that, Erik Zenhausern, director of acquisition and retention at Newsday, said a company needs four things:

1. Data and a data structure.
2. Analysts to interpret the data.
3. Marketing manager.
4. Creative services team.

Overall, it is more of a market-based process. “It’s really more of a philosophical change than a product change,” Zenhausern said.

INMA members can access the recorded Webinar for free here.

For more insight into how media companies are using data and predictive analytics, join INMA in London for Big Data for Media Week February 20-24.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .