Interview with Tor Jacobsen, SVP consumer marketing and revenue at Schibsted Media
Tor, Schibsted runs a number of newspapers and websites in two countries, nationally and regionally. Do they all have a paid content model?
All of our websites have paid content, our 16 newspapers also now lately Omni. We have now a kind of freemium model all over our sites, the share of the premium content varies from 10 to 15 percent paid content on some websites up to 50+ percent for other publishers.
But not all websites have a pure freemium model, do they?
The freemium or plus model is the base for all our websites, but for some brands like „Aftenposten“ we have put a metered model on top, so it´s kind of a hybrid model.
What are the reasons for following different strategies?
We simply think that the brands are different so different things work out. With our tabloids „VG“ and „Aftonbladet“ we have two big brands with a huge reach but still in a competitive market. In view of this we have a much lower share of paid content. The more regional or local the brands the easier it is to increase the share of plus-articles. Looking at our hybrid model, comparing the sales coming from plus-articles vs meter, we see that the plus-articles are taking more and more share of the sales. The plus sales often are like 5 to 10 times bigger than the sales from the meter.
If there is no one fits all solution what would you say are basic rules publishers can follow or should consider to find the right subscription models?
I think it is important to pay attention to find the right balance between the number of paid and the number of free articles. At „Aftenposten“ for example as the newspaper with the most subscribers in Norway we have increased the share of paid articles up to 35 percent where we see the limit. With a higher share we are expecting less people entering the website which means less conversions at the same time. This limit is different from brand to brand. It rather makes sense to set revenue goals and to evaluate them constantly than blindly increasing the number of paid content just because you set a goal before. At Schibsted we have set revenue targets and subscription targets for all our brands.
Measured by the number of digital subscriptions with 800.000 (Q4/19) Schibsted is one of the most successful media companies selling subscriptions in Europe. For 2020 you gave out the goal to reach a revenue of 1 Billion NOKs — will you reach it?
We sold out a couple of newspapers last year so that non-adjusted we’ve counted around 800.000 subscriptions now. The number has increased by 50.000 since Covid-19. Including print + digital and only print subscribers the number is by 1.3 Million Subscribers in Sweden and Norway. The development makes us still feel optimistic to reach that revenue goal till the end of the year. We gave out the goal in 2017 when we counted a revenue of 500 Million NOKs.
Success will not only be due to your products, but also to the users, who are significantly more willing to pay for digital journalism than in other countries. Why is that?
The biggest driver is of course we are doing a fantastic job. But seriously: This can have many drivers. One, I guess, may be historical. People from Sweden and Norway have always paid more for journalism and subscribed to several newspapers. A completely different culture has emerged from this. An additional factor might be that in general in Sweden as well as in Norway we see very digitized markets. Mobile as well as internet penetration came very fast. And of course we must not forget the financial situation. People are more likely to afford paying for information. Last but not least: Compared to for instance English and German speaking markets Sweden and Norway are probably less competitive with a high penetration of paid content in media.
Developing a working subscription model means to go through different phases or rather focusing on different aspects — increasing the number of users, retention or optimizing the revenue stream. In what phase do you see your company?
Focusing on different aspects sounds more appropriate, indeed. Of course you want to increase the number of subscribers while you are focusing the optimization of revenue. For us focusing on the optimization of price and packaging currently has priority. This means we want to bring forward the right product packages for different audiences — like basic packages, premium packages, special sport or finance packages.
De-bundling and fragmentation does make it more complicated for users, doesn’t it?
De-bundling after you bundled the products sounds weird but it is important to understand that on the consumer’s side the needs and the demand is constantly changing. Being flexible is the most important thing. Today using data allows us to analyze our knowledge about users’ needs deeper than in the last years. But you are right. You have to be careful not getting messy.
You just said you increased the number of subscribers by 50.000 users since the outbreak of Covid-19. These users apparently came due to a special purpose. What do you do for retention now?
First of all we are trying to maximize the engagement by showing them more content that they could be interested in. During the crisis for example we have included an everyday home training with a personal home trainer as well as adding cooking courses and even language lessons. We have trustworthy brands for many of our users, which is why they often trust us to include new features and products.
How does data help you to learn more about the users?
Data is getting more important and we care a lot about analyzing it. But there is still much more potential we are not using yet. Where we are using it intensively is at Aftenpostens frontpage for example. Of course not all the content but a but a significant number is driven by an algorithm.
Where do you use data regarding paid content?
At Svenska Dagbladet we are currently testing a data model which helps the newsroom to decide if an article should be paid or free. “The Oracle”, how it is called, only gives advice via Slack messenger, and even if the members of the newsroom might disagree they at least debate about an article again. It is a huge help in daily business with surprising results.
We’ve learned from Oracle that even information that users can get for free anywhere else brings them to convert in our subscription because they want to read the articles from our journalists. People in the media business tend to see their own content as raw material and underestimate the loyalty of some users at the same time. It’s important to understand that emotions are an important driver for people making decisions. Data models like this helps us to fear competitive situations less.
So how does it work in detail?
When we put an article online for free this data model is analysing the momentum and the user’s behaviour. When it recognises a significant rate of impressions or engagement — that’s the main driver for the software among other aspects — Oracle gives a signal and the newsroom is evaluating this article again.
What impressed me was your churn rate about 9 to 10 percent per month. That would be a dream coming true for German publishers. But you don’t seem satisfied.
You bring up an interesting topic we are still discussing internally. You might have a point here. It is a high churn if you compare it to Spotify or if you compare it to HBO or Netflix or even to other markets like selling cell phones or telecommunication subscriptions. But is it really a problem? As long as the people come back to us for finding information and some day they convert into a subscription our acquisition costs are like zero. So the easiest way to reduce the churn would be to stop every external campaigning like marketing or price campaigning. On the other hand this would mean to be more defensive on the sales side and it might not be smart to do. What would be good in the long run? We haven’t yet figured out an answer for us.
In the subscription business there is a rule that the first 100 days of subscriptions are the most important in users lifetime value. What happens within these 100 days?
This period describes the critical phase where the user decides to become a loyal subscriber or not. Within the first 100 days there is also a lot of campaigning which also means a high campaigning churn because users didn’t become a subscriber due to a real purpose but due to a good offer for them. Whatever: In our data we see that the churn rate for users after a hundred days goes drastically down. If you only look at this type of users it is around 3 to 5 percent.
The Interview was recorded in Summer 2020 and first published in Medieninsider‘s INSIGHT #1 about Paid Content and Subscriptions