“Translating Public Service Delegation into Algorithms”

Mr. Gaddum, Mr. Grün, more and more public service formats are being produced specifically for social networks. How dependent are you on platforms like YouTube, Instagram or TikTok?

Eckhart Gadom: Of course, we want to promote our own platform above all else. This is the core of our business. But where we don’t reach the audience directly in this way, we go where the people are. Our priority is to address the community as a whole. In return, we accept the presence on the third-party platform. Nobody is without sin – not even in the digital world. But apart from that, we not only operate in a rule-free zone there, but we also give ourselves a clear set of rules on external platforms.

for example?

Arrival: And our journalistic principles like balance, inclusivity, etc. apply there as well…plus our loyalty doesn’t belong to the platforms, but rather to the audience. We are guided by their expectations. It’s no different on TV. And if the caravan moves, we will go with it. We’ve already said goodbye to Facebook here and there.

However, getting TikTok users into the ZDF media library is likely to be a difficult task.

Arrival: Rusty bread, of course – especially since TikTok is something completely different talking. Convincing people to switch platforms is a challenge. But look at funk, for example. This is a great example of how to occupy a public place in a commercial setting. And now this content also contributes to the success of ZDFmediathek. Many, especially young people, do not know what we offer them. Think thousands of documentaries. And beyond the content – we also want to show that we treat data and algorithms differently – more transparent and more public. We want to be the “best” in this field as well.

What do you know about big platform algorithms?

Andrew Green: Experts who work for Netflix, Amazon or YouTube participate in big conferences and publish a lot. We now also know a lot about the Meta. Concepts, but also basic strategies through technological implementation are known. We are in an abusive discussion with Google in particular. But this is of course common knowledge among experts, for users the algorithms of all platforms are a black box. In short: we have similar problems but we deal with them differently.

Arrival: For us, it should be about translating public service authorization into algorithms.

What does this translation look like?

Green color: Our job is to actively get people out of the bubble and get them excited about new topics. To do this, we develop our algorithms ourselves, and we have to constantly improve them and correlate them with viewership and other success metrics. This is a challenging task involving editorial teams, developers, users, supervisory bodies, and content partners. It is important to be transparent and disclose laws and regulations. Just take a look at ZDFinfo’s start page in the media library, which is now 100 percent automated and thus very successful.

“But we firmly believe that diversity and serendipity will pay off in the medium term — for everyone.”
Eckhart Gadom

It’s always been said on TV that entertainment shows are important in order to get the audience’s interest in, for example, “Hewitt Magazine.” Is there such a thing as audience streaming in a media library?

Green color: Of course, audience flow is also a very important variable for us. Certainly, if people want to overexpose a show, we should allow them to do so. On the other hand, we also want to draw your attention to world events, for example through a link to a live broadcast or a recommendation for an information format that may be appropriate at this point. We take our users very seriously, but also our mission. Setting this balance in an arithmetic way is a very important task.

Arrival: We’ve done some AB testing, i.e. running different variants, and asked ourselves what actually happens when I show a “heute magazine” to a user who has viewed the “heute show”. The experiences are positive. We can’t be certain that success only comes when we keep people in their bubbles. Dealing with algorithms is showing a growth in size and an exponential increase in various uses. Half of the entire media library content is consumed every 14 days via algorithms and automation. No editor can build it!

Green color: Lots of niche content is also seen. The content caters to users through recommendations. Of course we are not satisfied yet, but we are looking at how to develop it further. However, we firmly believe that in the medium term diversity and serendipity will pay off – for everyone.

How important is it to enroll users in the media library?

Arrival: For us, the principle applies that all content must also be accessible without registration. Of course, we invite all users to register with us as this allows us to learn more about them and serve them better. We can often convince users to sign up via age verification. Another driver does smooth viewing, i.e. start playing the content on the cell phone and continue to look at the same place on the tablet in the evening.

Green color: We currently have around 3.7 million registered users. Five to ten percent of these people’s daily use can be returned. This is not a little, but not a lot either. We want to grow significantly by offering tangible added value.

How is it supposed to work?

Arrival: To do this, you have to do a lot at the same time: push personalization and automation significantly as described – this is the answer to the individualization of user interests. At the same time, the structure of research and presentation must be understood. Attractive visuals are needed – we can currently improve this a lot. Together with ARD, we are expanding the permeability of our standalone media libraries – by linking search systems, cross-playability and recommendations, and cross-login – that will create a large public world for users.

To what extent does the media library contribute to renewal today?

Arrival: Our average age at the moment is 48, so at least we have a foot in Bab al-Shabab. Our goal is to become an attractive platform for people in their 30s and 40s. We are working to achieve this with new content, but also with the further development of our front-end, usability and meaningful arrangement structure for the ZDF media library. The latter is often underestimated.

Not afraid of Netflix and all the other competitors?

Arrival: Look at the market calmly and confidently. There are key features that distinguish us from classic streaming platforms: we can do “live”. We know “events”. We see “diversity” as a strength. The way we handle data sets us apart from the rest. Line TV connectivity and joint planning and production operations remain strong competitive factors for us.

Mr. Gadom, Mr. Gron, thank you very much for the interview.

Leave a Comment