Driving the customer subscription lifecycle is key to a reader revenue model. And this begins with experimentation.

At the INMA Media Subscriptions Town Hall on Tuesday, a panel or three news media executives shared their experimentation stories and how — through newsletters, paywall changes, better targeting female readers, a Web site audit, and premium content changes — they got closer to their reader revenue goals.

El País, Spain

Angélica Domínguez, customer strategy director at El País, said the company has evolved from a registration wall in mid-2019 to a subscription model in May 2020, then to a subscription app and loyalty programme in 2021.

“We have a mixed model,” Domínguez told INMA members. “We have the signwall, and we have implemented a metered paywall with 10 articles.”

This has resulted in 2.6 million new registered users, bringing the total to nearly four million; 125,000 of those are digital-only. Although this represents big growth, recently that growth has slowed.

Like the other media companies featured at the Town Hall, El País took part in the Google News Initiative (GNI) Subscription Labs Europe in 2021, reaching several main conclusions. They identified their strengths:

• Huge brand awareness.

• Good quantity and quality of traffic, both national and international.

• Low churn rate, especially involuntary churn.

• Good consolidation of data.

They also identified El País weak areas where they wanted to improve:

• Low meter stop rate, which translated into a decrease in organic acquisition.

• Low loyal female and young audience.

“El País has begun to use promotions to attract new readers to become subscribers,” Domínguez said. “That was having a negative impact on the ARPU and on the churn rate.”

With the GNI conclusions in mind, El País undertook several experiments working with FT Strategies:

• Tightening the paywall. This experiment is still underway.

• Increasing female engagement.

• Dynamic offerings.

The experiment to reach out to and increase female readership started with a hypothesis: The team believed it was not only about producing the right content for women but also placing that content in prime positions. 

The team used real-time data to identify articles of interest to women, selected the highest-ranking of that content, and promoted it on the home page. The test consisted of launching two different home pages, one highlighting content aimed at female readers.

“The results were positive,” Domínguez said. “The global CTR of the two home pages was similar, however, the female CTR in the blog of the home page with the promoted articles was 275% higher.”

The next experiment was to increase acquisition through intelligent offers. The team believed the use of promotions would increase conversions but could negatively impact churn and ARPU. Therefore, they needed to show offers to readers in a controlled way depending on their propensity to subscribe, thereby increasing their lifetime value.

The approach was to consider user profiles when offering promotions. This was tested by launching different offers to different segments of users. Domínguez said the results were very good in terms of lifetime value, increasing LTV by 34% in the first phase of the experiment and 25% in the second phase. However, the conversion rates were low.

“Given the results, it was a good learning for us,” she said. “We are using the results and working on extending this approach to a wider audience.”

The Courier, United Kingdom

Graham McDougall, head of audience management at The Courier (part of DC Thomson), said his company is currently about 18 months into a five-year plan to make its newspapers profitable through digital revenue.

The first step the team did was a Web site audit. Many of those findings came as no surprise, such as DC Thomson relying too heavily on social traffic.

“We knew we didn’t get enough users to the subscriptions page, and we knew our involuntary churn was high, but being able to rank and prioritise these gave us real focus,” McDougall told INMA members.

The team came up with ideas for solutions to close those gaps, both internally and from FT Strategies, which worked with the team in brainstorming sessions. From there, they went into an experimentation phase with some of those ideas.

“We intentionally set up our planning to focus improvements in each of the areas of acquisition, engagement, and retention, and set outcomes for business improvement in these areas,” McDougall said. 

• Acquisition: Increase the meter stop rate.

• Engagement: Reduce dependency on social traffic and increase quality, direct traffic.

• Retention: Reduce involuntary churn.

For the acquisition experiment, the team decided to increase the amount of premium content that was offered behind The Courier paywall. A weekly meeting between editorial and sales/marketing shared data on what worked and what didn’t, and the team had to agree on the conditions under which content would be made premium.

“The impact of these sessions was fairly immediate,” McDougall shared. “The average sales in the eight-week experiment period was 182% higher than our sales were before that. But it didn’t actually stop at that point because we knew we were onto something.”

This success inspired the team to innovate other ideas on how to utilise premium content to stop traffic that avoided the meter. Sales continued to rise as the result of these optimisations.

For the engagement experiment, the team focused on increasing traffic from newsletters. They decided to create three new newsletters for the start of the football season, with a hypothesis that would increase engagement metrics. 

From this experiment, The Courier no longer relies on traffic from social media as the dominant source. While the change in traffic percentage from the e-mail newsletters was fairly small, that traffic was higher quality and had better engagement. They became some of The Courier’s best-performing newsletters in terms of open rates and CTR. 

Not only that, but the experiment showed the team that they could create newsletters quickly. By launching several newsletters at the same time, they discovered the areas where they could best succeed — which was sometimes not what they expected. For example, rather than the best-performing newsletters being about major football clubs, they were more niche-focused newsletters about smaller local football.

The retention experiment to reduce involuntary churn, which at the time accounted for about 25% of The Courier’s churn. The team extended the past-due period from six to 14 days to increase the percentage of payments made and improve payment recovery rate. This required only a simple setting change in the payment tool, Braintree.

As a result, the payment recovery rate increased from 24% to 41%. 

“We are bringing more subscribers back every day to our subscriber base,” McDougall reported.

Ruhr Nachrichten, Germany

The North Star goal at Ruhr Nachrichten is to finance the newsroom entirely through digital reader revenue by 2025.

Philipp Ostrop, head of product development at the company, said this challenge means sticking to its journalistic mission without print, and to proceed as if print would no longer exist. 

“In order to achieve that, we need 111,111 Plus subscriptions,” he told INMA members, referring to the paid premium subscription offering at Ruhr Nachrichten. Currently, there are only 6,000 Plus subscribers, so they have quite a journey ahead. To achieve this goal, they must double subscriptions each year.

To achieve this, the team set up three experiments — or, as Ostrop said, one experiment and two data science challenges.

• Project Paywall. Current paywall offers are too complicated, and the team aims to change a variety of paywall design features to make it more simple and reduce friction.

• RFV Sweet Spot. This entails an analysis of the RFV score at the point of becoming a full-price Plus customer, to identify a correlation and tipping point. This will be used to contact readers at the right time for a subscription offer.

• QR-ious. The team wants to create a new and richer engagement metric. As an alternative to daily active subscribers, the best-performing content would be identified based on media time and quality read scores.

For Project Paywall, it was decided that any trial offers should be at least three months — instead of the previous 30 days — because it takes that long to build a habit. This was launched only one week ago, but Ostrop shared early results showing a huge uplift in conversion rates, almost doubling.

“This version is now our new standard,” he said. “We want to move further with experiments in a higher velocity and also do real A/B tests.”

In the retention point data challenge, the team analysed a group of 2,000 customers with a three-month trial period. Of those, 500 remained active and 1,500 churned. They built segments to categorise active subscribers and drive campaigns to maximise retention.

“The retention point of our subscribers is at an RFV score of 44, so in other words our subscribers need to have a score of 44 in order not to churn when we move them to full price,” Ostrop said.

They are now in the step of segmenting those groups so that they can take action on offering campaigns to different segments.

Lastly, the team wanted to create a new, richer engagement metric for Ruhr Nachrichten content. To date, the newsroom has only considered a single KPI: daily active subscribers. However, that metric has weaknesses, such as only showing the number of subscribers that have clicked on an article.

“That’s not enough to face our RFV challenge,” Ostrop said. “To really measure engagement, we need something else. That’s why we are looking at new KPIs that better reflect the engagement and have a better effect on building a habit.”

The team has started to calculate other metrics such as time spent on the content and scores around quality reading.

“Those are our three main points of experiments and focus in the past week,” Ostrop concluded. “Making the paywall easier, finding the retention point of our subscribers, and then from that trying to find a richer engagement metric for our content.”

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