Pact with the devil

Henning Eichler’s latest study, »Journalism in social networks. ARD and ZDF under the spell of the algorithms,« was conducted on behalf of the Otto Brenner Foundation and examines the conflict between public value and platforming. In an interview, its author explains how algorithmic methods of working and platform logics influence journalistic content. Eichler – a radio journalist at Hessischer Rundfunk and Deputy Professor of Media Sciences and Digital Journalism at Hochschule RheinMain – calls for greater transparency from the operators of advertising platforms and a code of digital ethics for social media editorial offices.

Interviewer: Wolfgang Scheidt

Mr. Eichler, social networks are absolutely essential for ARD and ZDF. In your study, conducted in late 2021, you examined 751 journalistic items, a quarter of which were offered exclusively on advertising platforms such as Facebook, Instagram, Twitter, and YouTube. How does social media journalism differ from content broadcast in linear form?

Henning Eichler: One difference is the principle of communication: Radio and TV are based on classic one-to-many communication, i.e. one broadcaster or communicator spreads content to a large number of recipients. Social networks, on the other hand, work on the many-to-many principle: Communities exchange content, interact, and communicate with one other. As a result, the route of dissemination is no longer from one institution to a large group of users – content passes back and forth between communities and groups of users. Secondly, linear and non-linear editorial offices work very differently. Platform content is created specifically in order to achieve the greatest possible reach. Content cannot be simply transferred from the linear world into the digital world of YouTube or Spotify. Instead, journalistic content needs to be optimized for specific platforms during the creation process. First, the concept, target audiences, and age groups are defined, in order to select the appropriate platform for the intended peer group. Each item of journalistic content is designed precisely for this platform: How long can the item be? Items on TikTok are between 30 seconds and one minute long; YouTube allows a documentary of up to 20 minutes. The form of presentation, the tonality, graphical design elements, whether the reporter is visible on screen or even part of the action – all these design elements need to be considered and implemented for the platform used. Each platform has its own rules, which decide whether journalistic content works or not.

The recommendation algorithms of commercial platforms favor content that is emotional, polarizing, and brief – complex, in-depth, and balanced content often slips under the radar. How does this impact the item length, tonality, dramaturgy, and topic selection of »platformized« content?

Henning Eichler: One example is the video format »Deutschland3000,« which is produced by public service broadcasters for Instagram and Facebook. Its host Eva Schulz does not always appear as a journalist or presenter with professional distance, but instead takes a clear position. This is because the editorial office has recognized that items with strong opinions or clear positions encourage users to interact more on platforms. Greater interaction means that the platform algorithm prioritizes the content and continues to play it. Interaction thus leads to greater distribution and reach. As a result, an editorial office is well advised to design content in a way that makes interaction easier and triggers agreement or disagreement. In line with the logic of platforms, this is more effective than reporting that is balanced and distanced from a journalistic point of view. This shows how an editorial office not only recognizes the rules of platforms, but applies them to the way it works.

If advertising platforms determine which content is displayed to which users and in which context individual items are incorporated, what dangers does this present?

Henning Eichler: The fundamental problem is that those creating content for public service broadcasters are unable to influence to whom content is shown. As soon as an item is produced and placed on the platform, it is out of the producer’s hands – the algorithm decides. The editorial office thus relinquishes some of its autonomy to the platform. This becomes dangerous when editorial offices use such networks for their entire distribution. It is important to always also take an alternative, technically independent approach from the social media, just as public service media do by using their media libraries and their own websites on »funk,« the content network of ARD and ZDF. Of course, the offerings there do not receive as much attention as they do on the social networks. That makes it all the more crucial that journalists consider carefully how strictly to follow the rules of platformization, how much to be guided by them, and how they can ensure that they maintain and guarantee journalistic standards. In the everyday work of editorial offices, this means that journalists need to constantly weigh up, day in, day out, the extent to which content is guided by the rules of the platform economy and how tightly to hold on to the framework of journalistic values. Every journalist who produces content for social networks faces this dilemma. There is no right answer: Every individual needs to constantly scrutinize their own work and reflect on the circumstances.

In your study »Journalism in social networks,« you examine the fundamental conflict between public value and platformization. When social media content is funded by license payers, to what extent is it guided by algorithmic methods and conventions of advertising platforms?

Henning Eichler: In the surveys conducted for my study, one editorial office explicitly said that Facebook users pay little attention to content that is more complex, subtle or multi-layered. The Wirecard scandal, for example, had no chance of gaining a wide reach on Facebook. As a result, the respondent claimed, editorial offices no longer offer this kind of content. On TikTok, environmental and nature protection topics receive little acceptance among users. One of the findings of my interviews was therefore that, due to the lack of resonance generated, editorial offices no longer offer certain content on social networks, even though it would be relevant to society and important journalistically.

The 18 staff from social media editorial offices and editorial office managers you interviewed felt that their journalistic work was significantly impaired. How strong is the influence of platform logics on editorial decisions and journalistic actions?

Henning Eichler: Platform logics have a clear influence on editorial work at various levels. The first is the development and design of new ideas for formats. Editorial offices primarily ask themselves two questions: what chance a format has to achieve an acceptable reach in which social network, and which platform they should produce it for. Benchmarks for how high a quantitative reach could and should be are defined during the development process. Every step in the editorial process is linked to the currencies of success on platforms: choice of topic and form of presentation. All editorial offices that work for social networks regularly analyze metrics and analytics from platforms. This platform economy is accepted by the editorial offices and incorporated into their day-to-day work. Data used includes the interaction rate, the dwell time – how long users follow an item for and when they click away –, click rates, and the amount of reaction and comment. All these currencies have been invented and developed by the platforms themselves for their platform economy business model. The editorial offices adopt the platform logic in its entirety in their work, allowing it to influence the way editorial items are assessed. It is normal for platform analytics to be discussed and analyzed in editorial office meetings. The journalists observe in great detail why certain content works well and what was right from the point of view of the platform economy – or vice versa. That means that the platform currencies directly influence the work of the editorial office; journalistic work is evaluated and analyzed with the platform logic in mind. Needless to say, editorial offices deal with this in different ways. For some, it is the essential foundation on whose basis decisions are made. The »funk« editorial office, for example, cancels formats if they do not attract certain figures. Other editorial offices do not consider the figures quite so crucial, instead trusting their own value framework, which they see as more important than the platform logics. Both sides exist, but no editorial office can ignore the data and figures of platforms entirely – they are present in every editorial decision.

A ›code of digital ethics‹ could clearly define journalistic work on advertising platforms. What would it need to look like?

Henning Eichler: Firstly, all stakeholders would need to take seriously, accept, and discuss the problems of the platform economy for public service content. The interviews with journalists who work for social networks show that the topic is an issue, a concern and sometimes a burden for them in their everyday work, despite being little noticed by the public. All media institutions need to publicly address the fundamental conflict in a debate in civil society. A second priority must be to regulate platforms or at least to ensure that public service content for the public good is not disadvantaged by the algorithm. An even better solution would be regulation by the European Union in collaboration with the EBU (European Broadcasting Union), configuring algorithms to favor high-quality public service content for the greater good on the platforms – all implemented by a supervisory and regulatory body. This would allow content that is properly researched and relevant to society to find more users on the platforms. At the same time, editorial offices would no longer need to tailor their work to the platforms as much, but instead could be sure that their public service content would reach a large number of users through special treatment by the algorithm. In addition, public service institutions should do much more to build their own technical infrastructures. The collaboration between ARD and ZDF on their media libraries and more self-developed apps are a useful way to offer more content that works on a different algorithm logic. The algorithmic selection can prioritize content that is well-researched and especially balanced, report on background, highlight new perspectives, and display a transparent structure – all quality criteria that are familiar from conventional journalism. Of course, this can only work on providers’ own platforms and digital infrastructures. Public service media need to invest a great deal more money and resources into developing projects like this.

Your study generated a huge response when it was published. Tanja Hüther, Head of the ARD Distribution Board, for example, considers it essential that public service media use only the »good side of technology.« Surely a refusal to use commercially oriented platforms such as YouTube, Instagram, TikTok, etc., is counterproductive in terms of reach?

Henning Eichler: If Ms. Hüther is referring to how public service media uses platform environments, this kind of separation is impossible. Using only the good side of technology and leaving out the bad would never work there. The two sides exist together. To some extent, distributing content on social networks means entering into a pact with the devil. On the one hand, it enables a large reach and access to target audiences that cannot be reached elsewhere. On the other, it demands a willingness to play along with certain rules and logics of the networks. Large quantities of data are gathered on all users, and there is no transparency on which data is gathered, how it is used, and which user and usage data is provided to the public service media – that is all part of the platform economy. How exactly the algorithms work and which criteria the recommendation systems are based on remain unclear. Using social networks for distribution means accepting all this. That is the other side of the coin.

So what does Ms. Hüther mean?

Hennig Eichler: If Ms. Hüther is referring to the institutions’ own technical infrastructures such as their media libraries and proprietary systems, it is certainly possible to use only the good side of technology. To do this, they would have to set up their own algorithmic recommendation systems that are guided by values and work based on ethical principles and value frameworks. However, this infrastructure needs to be advanced and established first, before users will be willing to spend more intensive time in a media library, for example, and to reduce or entirely eliminate their use of YouTube. At the moment, media libraries simply cannot keep up with the functions of social networks. Furthermore, there has not yet been satisfactory success in shifting users from social networks to public service platforms. This year, ARD set itself the strategic goal of tempting users in the social networks with brief items and excerpts in order to attract them to their own platforms, such as their media library. Yet despite numerous attempts and experiments, the campaign has not been a resounding success. If institutions are to be able to use only the good side of technology and reach young people, users need to be encouraged much more effectively to switch from the commercial platforms to public service platforms.

Does it make sense for public service media to bring on board influencers who already have reach, rather than establishing their own reach?

Henning Eichler: Influencers are strict followers of platform logic, achieving their enormous reach and number of users by understanding and following the rules of the game so effectively. The question is, would influencers retain their great reach if they were to present more value- and quality-based content on the social networks? If successful influencers are only present in the ARD media library, that would be yet another environment in which users would have to follow them, and in which they are not offered the interactive and participative social media functions with which they are familiar from platforms like YouTube – options to discuss with the community, connect with others, give the editorial office feedback, or use tools like surveys and quizzes. The ARD media library does not offer functions like this and, until it does, users are unlikely to move away from the social networks they know and love.

Can methods of working in the social media sector be applied to the linear channels, so that they can work more efficiently and get young target audiences excited about public service television and radio?

Henning Eichler: Absolutely. For one thing, linear environments can adopt the much stronger user orientation of social networks to get closer to what users want and need. Some editorial offices and innovation departments are now developing prototypes for new formats and programs, samples of which are being presented to the target audience and discussed in focus groups. This user orientation is useful when developing formats, in the feedback process, and when developing programs further. In addition, it provides a great deal of potential when it comes to opportunities for the TV and radio audience to interact and participate. For example, community management makes it possible to create a feedback channel into the editorial office. Both of these points are suitable for linear formats with no need for significant adaptation. In linear environments in particular, it would be useful to be more attentive and agile in reacting to changes in users’ needs, in order to remain closer to the users and to adapt and further develop programs and services more quickly.

When it comes to media policy, the latest version of the State Media Treaty [Medienstaatsvertrag] is intended to give public service providers greater flexibility to play content on advertising platforms such as social media. Internationally, the EU Parliament has passed stricter rules for internet platforms. The Digital Services Act (DSA) is intended to put checks on hate speech and other illegal content online; the Digital Market Act (DMA) to curb the market power of large internet corporations. Will this reduce the dependence of ARD and ZDF on Meta, Google, Apple, and Amazon, and will advertising platforms reassess their fundamental business model?

Henning Eichler: No, the business model of platforms will not change at all. But platform operators will need to act with greater sensitivity on the European market. Public service media and everyone acting on social networks will remain dependent, as this dependence is structural and systemic. Neither the DSA and DMA nor the State Media Treaty will give the public service media greater autonomy. Despite this, the DSA sends an important signal in standing up to the large platforms. For the first time, Europe is attempting to create a new rulebook in order to regain the ability to take action. In recent years, media policy has merely reacted to dynamic developments in the platform economy. This new set of instruments could enable policymakers to actually fulfil their duty of supervision and regulation. Public service media have gained an opportunity to provide content and have it prioritized by the algorithms. The State Media Treaty states that platforms are not allowed to discriminate against any content. Conversely, this means that platforms are not actually allowed to prioritize using algorithms – which is, of course, absurd. According to the State Media Treaty, media institutions are responsible for supervising platforms. If supervisory bodies find that content has been discriminated against, they can demand that the platforms alter this immediately – although it is unclear whether this would really enable public service content to achieve greater reach. It also does little to change the fundamental asymmetry between global digital corporations and more regional media institutions. What really matters is whether Europe succeeds in installing an effective system of supervision, regulation and sanctions, in order to signal to platforms that Europe has stricter rules than the rest of the world.

What would need to change in order for the black box that is the algorithms to become more transparent and for journalistic work to have the same chance of distribution on the social networks?

Henning Eichler: Platforms would need to be forced not to use algorithms to disadvantage, or indeed advantage, content that is for the common good. This would allow high-quality content to reach the greatest possible number of people. From a technical perspective, a two-column function could be installed on platforms (for source, see Schwartmann et al. 2020): Users would be able to choose between a commercially oriented algorithm and one guided by the common good. For example, in the YouTube app, the general algorithm lists YouTube’s suggestions. With this new function, it would be possible to switch to the public value algorithm within YouTube and thus receive a different portfolio of content, oriented on quality, common good, diversity, and balance. This public value algorithm would need to be developed and supervised in collaboration with an independent organization.

What about platform transparency?

Henning Eichler: The DSA includes a demand to allow academia to access the algorithms and the way they work. Yet it does not specify how the data came about and how complete the data sets provided are. Giving journalistic organizations access to some of these data sets and algorithmic ways of working would also be useful, allowing editorial offices and other quality-oriented organizations to understand how algorithmic systems of recommendation work.

What role could journalism studies and communication science play in mediating between public value and platformization?

Henning Eichler: Science can provide food for thought when it comes to the idea of public value. The public service media take a rather superficial approach here, which does not cover every dimension of public value. Little attention is paid, in particular, to the debate in civil society on what public service media needs to provide, not least in commerciao media environments. I also see it as part of the role of academics here to isolate and explain the meaning of platformization and its consequences. These form an essential basis on which to make strategic decisions and carve out a clear position for public service media. It would also include thinking about a code of digital ethics – another field in which science could provide inspiration.

Will linear content eventually merge with digital content, or will the two worlds remain separate?

Henning Eichler: That depends a lot on the usage situation. When I am sitting on the sofa in the evening, I will use the media library, linear television, or a longer podcast on my smartphone. This relaxed situation is fundamentally different from a ride on the subway, when I would choose to get a quick overview of the news from Instagram. In that sense, I do not believe that the two worlds will align. Linear media will continue to decline in relevance in society and become less important in day-to-day routine media use, but they will not disappear completely, retaining a fixed position in people’s daily rituals. Questions surrounding format are more crucial: Which services do I use in which situation in my life? While cooking or ironing, for example, a user might have an hour and choose to enjoy a podcast or feature. Someone looking for a ten-minute news update might opt to get their briefing from a short news podcast or the news bar on their smartphone. In the future, the question will be not whether linear or non-linear, but about formats, needs, and usage situations.

The study »Journalism in social networks. ARD and ZDF under the spell of the algorithms?« is available to download for free at:

Translation: Sophie Costella

About the author

Wolfgang Scheidt (*1967) graduated with a master’s degree in communications, psychology and media law from Ludwig Maximilian University (Munich) in 1998. Since 2005, he has been a permanent OnAir Manager at the Seven.One Entertainment Group of ProSieben Sat.1 Media SE. As a freelance journalist he regularly writes about media topics for publications such as mebu live and Journalistik.


Schwartmann, Rolf; Hermann, Maximilian; Mühlenbeck, Robin L. (2020): Transparenz bei Medienintermediären. Edited by Medienanstalt Hamburg/Schleswig Holstein. Leipzig: Vistas.

About this article



This article is distributed under Creative Commons Atrribution 4.0 International (CC BY 4.0). You are free to share and redistribute the material in any medium or format. The licensor cannot revoke these freedoms as long as you follow the license terms. You must however give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits. More Information under


Interview with Henning Eichler: Pact with the devil. In: Journalism Research, Vol. 5 (3), 2022, pp. 299-308. DOI: 10.1453/2569-152X-32022-12701-en




First published online

December 2022