By Alejandro Serrano,
Mather Economics uses a data-driven churn analysis to help newspapers develop a targeted, effective and efficient retention strategy, according to Matthew Lulay, a director at Atlanta-based Mather.
“I know retention isn’t the most glamorous subject in the world,” Lulay told people at his Saturday, Feb. 25, presentation at the New England Newspaper and Press Association winter convention.
But Lulay said it is important to identify characteristics of a subscriber who might unsubscribe so that the newspaper can target that subscriber with a personalized retention strategy, rather than a holistic retention strategy applied to all of a newspaper’s subscribers.
“You have to be targeted, you have to have a personalized message. That is the key,” Lulay said.
To try to identify such characteristics, Mather analyzes several years’ worth of a paper’s transactional data, complaints, types of complaints, a subscriber’s payment history as well as digital data such as page views, stories read and amount of time spent by a subscriber.
Each element contributes to a comprehensive look of at-risk subscribers. For example, if a subscriber always pays his or her bill within a seven-day grace period for a year, and the following year he or she does so within a 15-day grace period, and the year after he or she does so within a 20-day period, then that subscriber is susceptible to churn, Lulay said.
Other information, such as the number of days a subscriber gets a newspaper delivered, is also indicative of churn behavior.
Once Mather analyzes such data, it can help a newspaper create an optimal retention campaign using the churn score produced by the analysis.
Lulay discussed different types of campaigns, including an incentive-based one and a dynamic messaging one.
Lulay used a case study of three different incentives — a $10 gift card, a smartphone charger, and a personalized postcard — to demonstrate the incentive-based campaign,
“Incentives matter across the board, regardless of what they are,” he said.
The effectiveness of the incentives varied, and Lulay noted that it is important to factor in the cost of the incentive into results. An almost $20 difference in cost between the charger and the postcard made the postcard a great deal.
Lulay said a dynamic messaging campaign consists of personalized messages to subscribers who are in danger of dropping their subscription. The first message is generic, whereas the third message is more direct and personal. Lulay said he likes dynamic messaging campaigns because they have practically no cost.
Mather’s system is like art and science: the science is the model and the art is how it is applied.
A well-defined churn model will predict churn rate, and that score has to be applied early and accurately to optimize its success, Lulay said.
“The key here is to start early,” he said.
A dozen people attended the workshop on “Circulation 2017: Moving toward data-driven retention” in the Boston Marriott Long Wharf hotel.
‘You have to be targeted, you have to have a personalized message. That is the key.’
—Matthew Lulay, Director
Mather Economics, Atlanta
‘Incentives matter across the board, regardless of what they are.’