Janis Kitzhofer from Axel Springer, and Bent Steeg Larsen and Mikkel Stampe Davidsen from Politiken share their best practices for editorial data strategy and management.
Axel Springer – Janis Kitzhofer
Janis Kitzhoferis a data evangelist with a diverse background in journalism and marketing. While working as a print editor at the Financial Times, Janis found that he had no insight into the behaviour and engagement patterns of the newspaper's readers. This prompted him to change his career path and move into marketing, where he gained 15 years of experience working with publishers.
In 2021, Janis returned to the news publishing business and is now Head of Editorial Analytics at Axel Springer. He bridges the gap between the data department and the newsroom, advocating for data-driven approaches to improve journalism and reader engagement. His mission is to make every journalist aware of the value of data and how it can support their work.
Janis' commitment to storytelling, data-driven journalism and empowering newsrooms has positioned him as a key expert in leveraging data in journalism.
Here are 5 key principles of editorial analytics from Janis:
1. Use empathy to build bridges between the data department and the newsroom
According to Janis, a creative approach and empathy are key to encouraging journalists to use data in their daily tasks. In his work, Janis draws on his experience as a journalist and follows the value of empathy to understand the newsroom's perspective. This helps him connect more closely with authors and editors on the team. In weekly workshops Janis listens to the journalist’s questions and needs, discusses the meaning and value of certain metrics, and most importantly talks about ways of how they can impact those metrics.
"If I showed up in a newsroom as a typical ‘tech guy’ or ‘data analyst,’ there would be a big gap between me and the journalists. So, I use fun and creative techniques to speak their language: Storytelling, interactive presentations, and in-practice workshops," Janis says.
"My main mission is for every journalist to know how looking at data can help them do their job, but I also want them to enjoy it."
2. Use metrics that are going to stick in people’s heads
For Janis, it's important to introduce metrics that mean something to those who use them and stick in their minds – which only happens if you have an immediate influence on them. "I negotiate between the different needs in the organisation and try to find a good middle ground. Goals are often tight to business metrics with limited informative value for the individual journalist," Janis explains.
"Let’s take visits for example. It’s an important metric for ad bookings and comparability with other publishers but it is not very tangible for authors while a bad read through rate is a clear call to action. Changing the structure, the form, the headline, etc. can impact the results significantly. And seeing these changes to have an immediate effect is what creates stickiness. I call it 'the power of live' in stark contrast to retrospective analyses that look at past events," explains Janis.
Equally important to him is connecting metrics and data with a story.
"My team has analysts and data scientists who are great at creating dashboards and data models. But we also have someone who can tell the story: me. All of that helps us teach people the importance of data and empower them to apply what we teach them".
3) Define the strategic pillars of your work
Axel Springer has three strategic pillars of work with data: educate, empower, and evolve."Educate" means bringing people closer to data and its meaning through reports and analysis, direct conversations and fireside chats with editorial teams around the globe and external partners. Janis sees knowledge sharing as an important part of learning best practices and mistakes that help the media industry.
"Empower” means giving people tools for data analysis and teaching them how they work so they can feel confident using them and gaining more insights.
And finally, "evolve" means providing a direction for the organisation. “In data analytics, you can look at many numbers, but you have to define which ones are the most important so as not to divert attention from the real work of writing and producing content,” Janis emphasises.
Data chart showing the article performance (image by Axel Springer)
Heatmap showing key metrics for Bild (image by Axel Springer)
4) Think about the long-term effects
When asked about his north star metrics, Janis replies that page views, visits, and subscriptions are important for Axel Springer as for any publisher. However, he points out that these metrics are interrelated and cannot be considered in isolation.
"Of course, some isolated metrics are popular with publishers because they drive the short-term business, are easy to understand and communicate. But they can also be counterproductive. For example, if you have subscription goals, you might think, 'Okay, I'll make 70% of my articles premium, put everything behind the paywall, and create more subscriptions.' But what results will such a decision bring in the long run? Visits and page views might go down – everything affects each other."
"Personally, I am in strong favor of engagement metrics like time on page, as well as read through and recirculation rate."
5) Share your best practices
As mentioned earlier, knowledge exchange is essential for Janis. He actively participates in conferences, summits, and talks to share his team’s learnings and takeaways.
“I talk a lot to other publishers and visit our internal teams across the world because I believe that this is my mission: inspire others with data”, he says.
Politiken - Bent Steeg Larsen and Mikkel Stampe Davidsen, Politiken, Denmark
Bent Steeg Larsen (L)is a Research Manager, and Mikkel Stampe Davidsen (R) is Head of Data at Politiken– a Danish newspaper with over 130 years of brand history.
Here are 5 key pieces of advice from Bent and Mikkel in the field of editorial analytics:
1) Build a data tool with a clear understanding of your customer's needs
In creating data dashboards, the main goal of Politiken's data team was to make it easier for key customers – journalists – to access and interpret what they see. For this reason, the data monitoring tools were integrated into the back-office system and the brand identity was maintained, including colours, look and feel.
Every time a journalist logs into the system or views an article through the company IP or VPN, he or she can see how it performs. All pages displayed internally have a data overlay that automatically pops up to show the most important metrics.
"It's important to us that journalists do not have to switch to another application to see the data about their work. It's displayed in their work environment all the time”.
2) Make data easily accessible
Another part of Politiken's data strategy also uses commonly accepted metrics that everyone in the newsroom understands. The most important part of the data display is a bar that shows the overall rating of the article based on several predefined parameters such as attention duration, read duration, bounce rate, and how many times the article was shared and saved.
Journalists can also see other easy-to-understand metrics, such as page views, average active attention time and conversion rate, defined as visits to the sales page for subscriptions and successful purchases. Another metric used by Politiken is the ratio of genders in an article to avoid gender bias.
As users scroll through the article, they can also see the average depth of the reader's visit, i.e. how many people read to a given point in the piece.
These examples of data visualisation make it easier for users to understand what they see and compare different content.
Most important metrics combined to form the overall article score (image by Politiken)
Scroll depth of an article as seen in the Politien data monitoring system (image by Politiken)
3) Make journalists the data-owners
Another important point in Politiken's data strategy was that the data should come from inside the newsroom, not outside. It is the newsroom that has ownership in this matter."The insights do not come from the analytics department - the journalists themselves track and analyse the data. You could say that our editor-in-chief owns the data, and when decisions are made based on editorial data, they come from him, not from the data team," Mikkel and Bent say.
"Of course, the data team is there to inform the business about KPIs and the overall "pulse" of the editorial team, but they are not there to give random advice to the editorial team. Their job is to see the bigger picture and discover patterns that can impact the overall state of the business," they add.
Politiken displays key metrics on large TV screens in newsrooms (image by Politiken)
4) Make data the talk of the newsroom
Mikkel and Bent admit that easy access to editorial data (Politiken also displays key metrics on large TV screens in newsrooms) has made data the "talk of the newsroom" at Politiken. In their view, this is a sign of successful data democratisation.
"Everyone sees these numbers, so people have something to refer to in their daily conversation. Every morning, editors look at the previous day's traffic and discuss: How many page views did we have? How many visits? How many sales? How does it compare to the benchmark numbers?" The journalists themselves also share engagement scores or page views, they say.
5) Keep your culture and values in mind
The data system at Politiken has not emerged without obstacles. There was some scepticism and reluctance when the data-informed approach was introduced to the newsroom several years ago. Crucially, however, journalists longed for feedback from their readers - and that's how Politiken treats the data.
"We could say it's a way to communicate and stay in touch with our readers. Data is not about judging your journalism; it's a feedback loop, a means of communication. And in our company culture, we respect that value very much."
They stress that you should never impose decisions that contradict the newsroom culture. For example, they do not compare individual author statistics to avoid a competitive environment.
"Our brand is very social, and our values are unity and community. And we have translated those values into our data strategy - we all work together for the same goal: to make good and reliable journalism," Mikkel concludes.
The text was originally published on Autentika's blog on June 19, 2023.