De Persgroep is a Belgian publishing company which owns media assets in Belgium, Denmark, and the Netherlands. They are one of Europe’s leading publishers and will host the INMA Media Innovation Week at their headquarters in Amsterdam this month.
De Persgroep has been using Content Insights for more than a year and Roy Wassink, their Insights Manager for the Netherlands, is our champion. In this use case, we talked to him about his experience in using Content Insights, as well as the way the De Persgroep staff use it. We also discussed what kind of value our application brings to them.
It’s important to note that data – and the implementation of that data – isn’t something which necessarily comes naturally to many editors in the newsroom. The challenge, therefore, is to inform the editorial teams as comprehensively as possible and to deliver reports which are as actionable and concrete as they can be. To this end, they’ve developed a unique way of using the application which addresses the individual requirements of their newsrooms.
In some of De Persgroep’s regional titles they are experimenting with a new organizational structure in which regional editors will be responsible for both print and online content in their respective areas. As some people may find this transition – and particularly the integration of digital into print – tricky, Roy and his team believe that the easiest way to introduce them to the possibilities (and benefits) an analytics-driven approach offers is to start off by using the system of reports he has set up.
Reports – “the answer generator”
Right now, as many as 70 different reports are circulating within the De Persgroep newsrooms in the Netherlands. These reports serve various purposes, from finding out which articles brought the audience to the websites the day before, to what kind of stories triggered loyal or engaged behaviors. This way Section Editors (who are each in charge of several Sections) can start their editorial meetings after reviewing insights from such reports. These documents need to keep them informed, without overwhelming them with too much information.
Roy explains that the reporting system is scaled on two levels.
- The first level reports include the so-called “volume list”. While these reports contain basic information such as the number of article reads, they expand beyond overly-simplified metrics, including information about read depth as well. This information helps them see whether people started reading the articles and, critically, how much of an article was read.
- The second level of reports are essentially a CPI list, but Roy and his colleagues have edited it slightly by taking away the volume statistics (ie the Exposure metrics). The reason for this? Reports are left with Loyalty and Engagement information on one side, and Exposure on the other. This way, editors gain more insight into which stories provide ‘volume’ (users and views) and which ‘value’ (engagement and loyalty). It makes editors aware of the fact that writing high traffic content or high-value premium content often asks for different stories – or a different approach to stories. Articles that score high on value are essential in converting visitors to subscriptions. Volume stories are, on the other hand, important for advertising.
Aside from daily reports, editors in higher positions, such as Cluster chiefs, News editors and Editors-in-chief, get weekly reports as well. Those reports help answer a wide range of questions. A weekly report answers the question of what type of articles encouraged loyalty and volume. These also explain which authors fostered a loyal audience. In addition, they also reveal which topics managed to attract more loyal readers. This extra information delivered to certain editors is a huge help in enabling them to make the right decisions on a day-to-day basis.
What is CPI?
Our Content Performance Indicator (CPI) is a one-of-a-kind algorithm that takes the evaluation of a website’s audience behavior to the next level. It’s the currency you need to explain how good your content is and how well it’s performing. We use a scored metrics system to evaluate the quality of your content – the opposite of the outdated and misleading single metrics systems used by many. CPI is made up of three behavioral models that separately value Exposure, Loyalty, and Engagement. CPI can be looked at both as a combination of all three CPIs (tailored to your business goals) or looked at separately.
Roy’s experience using Content Insights
Roy explains that he looks at Content Insights on three different levels.
- The first level includes the day-to-day, basic use of the application.
- The second originates from the questions he gets from other editors and his work here focuses on searching for answers about how their content performed.
- The third approach is one he utilizes when he needs to respond to more complex questions from editorial staff. These questions are usually along the lines of which kind of articles their departments should be writing more of, and which ones deserve less attention.
Fundamentally Roy sees the application as being something which can be used to speed up the changes in the newsroom and provide editorial staff with the information they need to make the best decisions over the next six months.
“Content Insights is a place that you can look into for useful information about what you did yesterday and what you did last month, and that can be an inspiration”
In comparing our application with other analytics tools, Roy’s observation is that other tools tend to focus on what is happening in a moment, without providing context for that moment. Content Insights, on the other hand, looks back to the past to provide answers to questions about why something occurred.
How Content Insights helped
Content Insights helped the editorial staff of one of the biggest news sites in the Netherlands, AD.nl, which has a daily readership in the region of two million people. The Editorial staff there wanted to know if it would be possible to lower the daily number of articles published, while remaining as successful in volume as they had been. Content Insights helped to answer the questions of what kind of articles were less effective by calculating the average reads on all the Sections and Sub-sections of the website as well as the average CPI. This then made it possible to identify articles that scored 20% or less of the average values and enabled them to make informed discussions about editorial strategy and where to place resources.
From vanity to inspiration: a different approach to analytics
Newsrooms need to start looking at analytics differently. It’s not about vanity (or at least it shouldn’t be) and it’s not about writing the story with the most clicks either. Data – at its best – should be a means to inspire the journalist, to show them what kind of articles their audience likes. And ‘liking’ shouldn’t only mean ‘what the audience clicks on’. It should also mean ‘what makes them return’.
To give us an example, Roy talked about politics. On many of De Persgroep’s higher-traffic sites, the topic isn’t all that popular: only a small portion of the audience will tend to read a story about a member of parliament’s opinion on a minister, or about what the opposition parties think about climate change. Switch the focus to how a new law might affect citizens, what they need to know about it and give real-world examples, and it’s a different matter. What’s the takeaway from this? Well, the political process in and of itself doesn’t make for nearly as successful an article as those about their practical consequences.
“We should spend much more time understanding what kinds of topics readers want to be reported on, but also the way they want journalists to write about them.”
From the data Roy sees that these are the types of stories with a high number of article reads and a high CPI. While it might be easy to assume that people are not interested in serious topics, this obviously isn’t true. It’s about the presentation of the information, about the tone and framing of the article. Roy says that media need to give people articles they can relate to and content that can help them in their daily life. As a result, he and his team at De Persgroep started making sure that editors have access to the kind of data that can give them such insights.
Identifying premium content and building subscribers
For De Persgroep staff, CPI is also useful when it comes to identifying premium content. The more usual way of looking at premium is through using ‘conversion rate per premium article’ as a metric. This does, however, have one big disadvantage.
“Loyal users of our products don’t make the decision to register or subscribe based on one article,” says Roy. “Rather, it’s the package over a longer period of time, culminating in the one moment in which they decide to take that next step.” This means that it’s important to understand what articles and what kind of articles make users more loyal and more engaged. CPI is a great indicator of that.
“If you’re looking at loyalty, in CPI you’re looking at user loyalty, and any content which those readers feel is worth reading is likely to have value to you in terms of conversions”
Through daily reports, the editorial teams get more insights in how to write ‘sticky’ content and his experience tells us that articles with a clear voice are more likely to convert people to subscribers than those without. Publishing original content tends to achieve the same result too, as do those which help readers in a practical way.
“The shift begins by talking about article reads, rather than page views”
Content Insights helps Roy figure out how De Persgroep staff spend time and what they are writing about – and whether those efforts are a good return on investment through readers’ attention time and number of reads. The ability to do input/output analysis is another useful function of Content Insights and something which isn’t possible on Google Analytics.
“It’s all about a cultural shift. The use of data might not be something that naturally to journalists and editors, and although vanity metrics are certainly useful in bringing data to the newsroom, they are not what it should be about.
The reports we use on a daily basis enable journalists to understand which and what kind of articles best suit the intention of the visitor: to understand what kind of articles best help De Persgroep as a publisher, both in volume (for advertising) and value (registration and subscriptions).”
The behavioral metrics system Content Insights created is important for De Persgroep in a few ways. It’s vital to keep their editors informed but not overwhelmed with information about content performance. Our application steps into the picture every time they discuss their audience loyalty, and it’s crucial when they’re identifying premium content. It empowers them to choose the right topics, and the application is a source of inspiration for journalists. Last but not least, the easy-to-use report system is crucial, and De Persgroep have editors developed their unique system of informing other editors about content performance.
“Are we there yet? No, of course not. But we’re on the important first step. With data comes awareness, with awareness comes discussion about what to write and how to write. And that’s what it’s all about.”
Our thanks to Roy Wassink for sharing his insights with us. Get in touch if you think Content Insights could help your newsroom or publication.
Marko Radojević also contributed to this article.