Analysing data around customer habits and behaviours is powering media companies’ efforts to fight churn.

Researchers and media leaders from Medill IMC Spiegel Research Center at Northwestern University, The Los Angeles Times, Corriere della Sera, and The Times of India shared how they are using data to understand churn and take steps to effectively reduce it during the recent Digital Subscriber Retention Master Class.

Multiple studies have been done in the United States and Europe to try to analyse multiple behaviours that are connected to churn, Greg Piechota, Readers First Initiative Lead, said. The relationship between a publisher and subscriber centers on frequency and the amount of content consumed.

“They usually show something similar around the world: It matters how long and how well we know each other,” Piechota said. “The longer someone is a subscriber, the higher probability they will stay a subscriber.”

Data has confirmed it’s easier to get someone to go from visiting once a month to four times a month, rather than get someone who visits 20 times a month to visit 25 times, Piechota said: “What matters in engagement is finding a threshold of engagement, not just maximising engagement.”

Medill IMC Spiegel Research Center

At the Medill IMC Spiegel Research Center at Northwestern University in the United States, Professor Edward C. Malthouse is research director, leading an initiative designed to make sure there are sustainable models for local news companies.

The team built a lot of churn models to study how different types of content, using different devices, and how people got to the Web sites affected the amount of clicking. They analysed how all these behaviours related to whether or not a user was going to continue paying. Then they found what they refer to as “the canary in the coal mine.”

“The best indicator is frequency or what we call regularity, so how often that customer is getting value from reading you,” Malthouse said. “If you get people to read more local news, they’re more likely to be retained.”

The type of content also mattered. Northwestern found the more people were reading local stories, the less likely they were to churn, Malthouse said: “If you’re getting some content that you’re not going to find anywhere else, then reading more of that content may make you more likely to stay.”

The Los Angeles Times

The Los Angeles Times has been testing behavioural data to better understand its relationship to retention and potentially even influence the behaviours themselves. Lauren DeLong, the company’s senior manager of marketing and technology, said there’s more to measuring time spent or visits per month: “Some things that we started to look at were overlapping data points.”

As DeLong’s team dug deeper into the data, they ran into dead ends, made some mistakes, and even ran tests that didn't go anywhere. Ultimately, they did have some findings that helped them with retention. They found that breadth of coverage was key.

Sports and business are two sections that typically draw in two very different types of subscribers. But if they subscribed to both, they stayed around longer. DeLong said there were similar findings with newsletter subscriptions: “We saw this also overlapped with subscribers where if they were only signed up for one newsletter over zero, they stayed a subscriber for three additional terms over a normal term length.”

Corriere della Sera

Corriere della Sera in Italy has evolved from focusing on an acquisition strategy to a customer centric strategy through data and personalisation in a relatively short amount of time. The company’s subscription journey started in 2016, but they really began seeing some traction by 2018 when its number of subscribers grew 220%.

“A part of this growth was, of course, related to the ability to reduce the churn rate of our subscribers,” Maria Sgromo, head of subscriptions for the brand’s parent company RCS Media Group, said.

Despite seeing massive growth, Corriere was able to cut its churn rate in half from 2018 to 2022. A new model developed with the data team is helping address and manage churn. Why were certain people more at risk of churn than others? The model tells them not only probability but the main reason why: price. Another main reason could also be engagement.

“In this case we understand engagement means recency, frequency, and volume. And based on this, we set some reengagement activities that are based on proposing more newsletters, proposing they download the app because we know people who use our app more are more engaged and come back frequently to check our news,” Sgromo said. “We develop specific content together with our journalists in order to provide some exclusive content to these kinds of subscribers.”

The Times of India

The Times of India (TOI) created a “plus” subscription and platform to give customers access to premium, paid content as a way to enrich their experience. Radhika Shukla, head of subscription growth, strategy, and analytics at The Times of India, said TOI+ has moved into the retention phase and is now focusing on the life cycle of its subscribers: “Not that we were not focusing on the subscriber life cycle management in our acquisition phase, but we are sharpening as we are growing.”

From there, it is all about keeping an eye on the metrics to track their retention rates.

“We ensure we track each of them separately and make efforts to work on each of them so that we can improve each of these metrics,” Shukla said, adding that of all the metrics they track, retention and renewals are the most important.

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