Can you rank the subdivisions of your news media outlet in terms of profitability of the customers who use them? Not the most used or the one brings in the most money — the division that attracts and appeals to the most valuable subscribers.

This approach forms the basis of what Peter Fader, professor of marketing at University of Pennsylvania’s Wharton School, is hoping to highlight — the importance of auditing your customer base.

During the recent INMA Subscriber Retention Master Class, part of the Readers First Initiative, Fader and media leaders from The Seattle Times, Hearst Newspapers, El País, and Schibsted Media shared how their companies are interpreting and acting on customer data.

Shifting the business strategy towards one that audits customers comes down to four core topics, Fader said:

• Determine how many customers you’re acquiring. Compare your analytics to the previous quarter, and the one before that, and the one before that, and figure out what is spawning this growth.
• Analyse customer retention — how long customers are staying with you and what is making them stay or leave.
• Determine how many value generating activities that your customers engage in. Any additional means of generating profit, be it ad revenue or additional products, should be analysed in tandem with customer interactions.
• Determine the amount of value that is generated from customer engagement, be it in the form of time or money.

“Acquisition, retention, repeat activity, and spend are the four things that I obsess over, that are at the heart of the audit,” Fader said. “They’re the four things that should be at the heart of every financial calculation as well, as that’s where our revenue comes from.”

How data analysis changed decision-making at The Seattle Times

While many media companies consider retention a goal, at The Seattle Times, it is a metric. Curtis Huber, senior director of circulation and audience revenue, said analysing data has affected how the company balances digital revenue vs. digital volume growth.

Since the goal is to keep overall subscriber volume fairly flat until it finishes this transition period and is in a digital-only space or a Sunday-only print product, it’s important to ensure that as the higher-priced, higher-profit print subscribers move to digital, the average rate remains flat.

It’s also about looking at such things as how individual sales are performing and how and how easy it is for subscribers to cancel — which research shows will make them more likely to return.

“That has led to a base that has grown to the size that even though we are keeping our retention rates flat, churn is not increasing,” Huber said. “For five years as we’ve grown, we have never seen our churn rates increase because of the way that we have managed it.”

However, with a growing base and falling traffic, Huber said it’s time to start pricing and testing more aggressive strategies with tenured customers’ renewal rates: “All of it is building into this need for us to grow volume and revenue — and one not at the expense of the other.”

Hearst Newspapers measures loyalty, churn risk

Like many media companies, Hearst Newspapers is looking for ways to battle a roller coaster of digital subscriber growth. Seeing a growth plateau in 2022 and retention decline in 2023, the company started focusing on the correlation between retention and engagement, Kelli Dakake, vice president of subscriber engagement and retention, said.

With that goal in mind, Hearst worked with Mather Economics to create a machine learning model that generates a score for each digital subscriber. The model measures the subscriber’s loyalty and risk by measuring 75 different variables. Those variables include behaviours like when they last paid, how long they’ve been subscribed and whether and how they use the newsletter. The score is from zero to 10 and is used to trigger certain actions.

Hearst found they have a large number of subscribers in the “at risk, low, and medium” range and a lot of very loyal subscribers with a lull in between.

“We use this distribution to target appropriate engagement strategies based on these subscriber segments,” Dakake said. “So we ran targeted e-mails and social media campaigns and it boosted engagement by about 7% to 11%. Interestingly though, on the other side, this test also showed us that our most loyal subscribers were actually negatively impacted when we tried to send these content campaigns.”

What the company concluded was loyal subscribers are already in a routine and cadence with content and don’t want other types of content pushed to them.

Audience segmentation at El País

Angélica Domínguez Martín, customer strategy director at El País owner Prisa Noticias, emphasised the importance of monitoring various stages of the customer lifecycle, including acquisition, onboarding, engagement, payment failures, cancellations, and win-back efforts.

Although El País manages around 50 different dashboards for specific KPIs, it does not have a single unified dashboard for all KPIs. One key metric is “Métrica Global,” which measures the performance of each premium article in terms of new acquisitions, time spent, pageviews, and subscriber reads.

El País leverages customer behaviour analysis to segment its audience, enhancing decision-making, and reducing churn. These segments are based on content preferences, such as political analysts, environmentally conscious readers, culture lovers, technology enthusiasts, explorers, and sports fanatics.

Each segment has specific actions tailored to their behaviour, such as hidden offers for promo seekers and preventive messages for those with failed payments. Studying the movement between these segments is also an important factor.

“We consider this as the key to measure the healthy evolution of our subscriber portfolio and to understand if our strategy is working correctly. What we see is how clients move from one segment to another over time,” Domínguez Martín says.

Schibsted leverages data to enhance the customer lifecycle

Tor Marius Espedal, head of customer base engagement at Norweigan publisher Schibsted Media, leads a cross-functional team that includes specialists in customer experience, data, analytics, and machine learning.

Collaboration between the insights and predictions team and the customer lifecycle team has been effective — and the two together function like a research and development team to figure out how to create actual value from the machine learning model.

“Looking at what the customer is engaging with in their paid subscription, can we see signals in terms of ‘What we can upsell? Can we get them to engage in other products that we have that they may not know about?’” Espedal said.

Zooming in on specific parts of the customer lifecycle has been really useful, Espedal said. For example, the team used its churn model to reduce cancellations in its automated upsale journeys by excluding customers above a certain churn score.

Schibsted also combined its content personalisation engine (called Curate) with its marketing automation tools to improve 1:1 personalisation capabilities. Curate can, for example, remove articles the person has already read from the front page and combine different article feeds to suggest more articles to read.

Since implementing it, Schibsted has seen much higher engagement with content and, interestingly, much lower churn in reaction to an e-mail reminder that a customer has a subscription — something required by Norweigan law.

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