How to Sell Democracy Online (Fast) (english)

November 2025 Melanie Weiser & Paulina Fröhlich Pablo Jost & Hannah Fecher November 2025How to Sell Democracy Online (Fast) 22 Foreword 3 Executive Summary 5 1. Introduction 8 2. Addressing Young People Digitally in Democracy: How to Sell Democracy Online (Fast) – and Why? 9 3. Subject of the Study 14 3.1 Instagram and TikTok: new platforms and formats for political communication 14 3.1.1 Platform architecture and platform-specific affordances 14 3.2 Research gap: our study 20 4. Study 21 4.1 What politicians send 21 4.1.1 Political actors examined 21 4.1.2 Content analysis 24 4.1.3 Success factors 34 4.2 What young people (don‘t) want 36 4.2.1 What the first five seconds reveal – a selection experiment 46 5. Recommendations for Action and Outlook 50 6. ”How to sell” in a nutshell 56 The Authors and the Team 58 Acknowledgment 61 Partner 62 Bibliography 63 List of Figures 73 Appendix 74 Imprint 77 22 Table of Contents November 2025How to Sell Democracy Online (Fast) 33 The perspectives of young people are often overlooked by politicians, even though their contribution is indispensable for a vibrant and sustainable democracy. As active democratic participants of today, whose attitudes and decisions will have a decisive impact on the democracy of tomorrow, young people should play a central role in democratic discourse. However, youth studies conducted in recent years paint a picture that needs to be taken seriously: many young people feel that their generation is not being given sufficient consideration by political decision-makers. Their view of the present and the future is also shaped by fears, such as the consequences of climate change, economic uncertainty, military conflicts and media polarisation. They see too few effective responses and too little accountability in politics in these areas. This has resulted in a loss of trust: the results of the 2025 federal election show that many young voters are disappointed and harbour fears (of social decline). They are turning away from the democratic centre and increasingly supporting opposition, radical and populist parties. Although the change in young people’s voting behaviour cannot be attributed solely to their use of social media, there are likely to be connections. The absorption of political content through journalistic – and thus usually editorially responsible – quality filters is increasingly giving way to algorithmically generated, unfiltered content from social media feeds, where populist and extremist actors have been particularly successful in placing their messages. This makes it all the more urgent to address young people in a more targeted manner and involve them more effectively in democratic dialogue. Our study shows: based on their own statements, young people view platforms such as Instagram and TikTok as being the most important places for them to come into contact with political content. These platforms are no longer merely places of entertainment, but also arenas for political information, opinion-forming and advocacy. In the daily environment in which young people move, short videos have become the dominant format: political content flickers across millions of mobile screens, mainly in the form of clips lasting a maximum of two minutes, and, whether consumed consciously or unconsciously, has a lasting impact on the political mindset of its recipients. The problem is that, despite growing efforts, moderate-democratic actors continue to be less effective in their communications than populist and radical forces. Democratic actors face the important task of learning: how to sell democracy online (fast). This is where political institutions and decision-makers bear a particular responsibility. They not only have to comprehend social media as an extension of traditional forms of communication, but also use it actively, attractively and strategically to reach young people. This means: content must be tailored to the target group, be understandable and visually appealing. Above all, however, the form of address must be credible and the core message of the clips must respond to the realities of the life and interests of the young generation. To this end, it is necessary to develop a deeper understanding of which styles and content appeal (or do not appeal) to young people on TikTok, Instagram and other platforms. Foreword November 2025How to Sell Democracy Online (Fast) 44 This study aims to contribute to this. It is intended as a guide for democratic actors looking to strengthen their social media presence and further develop their media literacy in order to offer young people attractive democratic information and identification opportunities. After all, at a time when democracies are coming under pressure in many places, communicating democratic values and attitudes is anything but a minor issue. This is not about mere click rates or short-term election advertising, but about building a vibrant democratic culture that also flourishes in the digital space and inspires enthusiasm. Dominic Schwickert Das Progressive Zentrum Regina von Görtz Bertelsmann Stiftung Marek Wallenfels Bertelsmann Stiftung Christiane von Websky Stiftung Mercator November 2025How to Sell Democracy Online (Fast) 55 Instagram and TikTok are the most important spaces for young people to engage with politics • When young people take to their smartphones, they use Instagram (81%) almost as often as messenger services (84%). TikTok (60%) is also used very frequently. Social media (74%) is, by far, the place where young people acquire their political information – even coming out ahead of family (58%), school (60%) and friends (54%). Newspapers and TV lag behind (46%). • 38 per cent specifically follow accounts owned by political parties or politicians, while 60 per cent follow political influencers. Half of young users state that they frequently view political content through algorithmically selected feeds. • 64 per cent of young people believe that social media platforms are a good place to reach their generation. Around half of them state that they appreciate the low-threshold contact options on social media (e.g. asking questions, hearing other perspectives). However, only 17 per cent actively participate in online discussions themselves, and only one in five likes or comments on political posts. This study consists of an AI-supported content analysis of some 31,000 short videos posted by political actors and influencers on Instagram and TikTok in the time ranging from June to December 2024, a representative online survey, as well as four focus groups comprising 27 young people between the ages of 16 and 27, along with a quantitative selection experiment in which young people were surveyed on their selection criteria for video features. Most videos revolve around the political process itself, with young people playing a subordinate role • Over one in three political videos on TikTok or Instagram in the second half of 2024 focused on actions undertaken by the government and administration (21%) or on elections (17%). • Just under a quarter of all videos contain a specific reference to young people or future generations. This share is lower (-5 percentage points) when it comes to the top trending topics mentioned above. When the topic of education is addressed (3%), which is rarely the case, explicit references to young people exceed the average (+9 percentage points). That being said, videos about education achieve a significantly lower reach (-17 percentage points). • Youth organisations of political parties in particu - lar establish a connection to young people and future generations (48%); political parties (21%) and politicians (20%) do so much less frequently; the AfD comes in last here (14%). • Videos that include references to young people generate greater reach and tend to attract more interest among young respondents. Executive Summary November 2025How to Sell Democracy Online (Fast) 66 Whilst acclaims prevail, it is the attacks that generate reach • Almost 70 per cent of all posts contain some form of acclaim. 35 per cent contain attacks on political opponents (ranging from criticism to insults). Self-praise is employed almost 20 percentage points more often on Instagram than on TikTok, while attacks occur 15 percentage points more frequently on TikTok. • At just under 90 per cent, Volt and SPD post acclaims most frequently, with the AfD (63%) coming in last. In terms of attacks, the picture is reversed, with the AfD accounting for the largest share at 73 per cent, followed by BSW (60%) and The Left Party (48%). The SPD (23%) and Volt (13%) attack the least. • Video clips containing attacks are viewed around 40 per cent more often on average than those without. That being said, ”views” do not equate to approval. An acclaim significantly reduces the level of reach (-12%). However, our surveys show that young people take a critical view of verbal attacks and reject the idea of disparaging others in short videos. Videos containing acclaims are not well received. Addressing migration increases reach, environmental and social issues do not • Videos concerning migration are viewed around 11 per cent more often on average. It should be noted that the attitude expressed in this regard was not recorded. Posts on elections (+8%) and videos without a political reference (+9%) also significantly increase reach. However, in an experiment involving young people and their preferences, only the former was well received, while videos without a political context or containing election references tended to be rejected. • By contrast, any focus on social policy (-7%), the environment (-18%) or education (-17%) had a detrimental effect. Selfies perform well in terms of reach, dancing and infographics do not • Videos are frequently recorded in selfie mode (17%); this occurs more than twice as often on TikTok than on Instagram. While political influencers record almost half of their videos this way, only 13 per cent of politicians do so. The overall evaluation demonstrates that videos in selfie mode have a positive effect in terms of views. • Infographics are most frequently employed by the Greens, The Left Party and Volt (9–11%), with the CDU/CSU and FDP employing them much less frequently (4%). Videos incorporating infographics achieve a lower reach on average. Based on the experiment, they do not make an impression on young people, even though half of them explicitly state that they want more figures and statistics. • Dancing appears in two per cent of all short political videos. Youth organisations (6%) and political influencers (4%) lead the way. However, dancing in political videos has a significantly negative effect (-45%) on young people: they swipe away. Calls to action: offline rather than online • All actors surveyed address the majority of their audience directly (e.g. using the informal or formal ”you”) – political influencers most often (66%); politicians least frequently (55%). While the form of address has no influence on views in the overall evaluation, it has a significantly positive effect among young people. • In just over two out of five videos, recipients are urged to take action. At 26 per cent, calls for online action (e.g. commenting or sharing) dominate, whilst having a negative impact on reach. This is followed by offline actions, such as calls for demonstrations (17%) and calls to vote (8%). These have a positive effect. Centre-left parties call for offline actions more often than centre-right parties. November 2025How to Sell Democracy Online (Fast) 77 • Calls to join parties (topping the list: youth organisations 8%) or donate (topping the list: political influencers 4%) are rare. Politicians are the most cautious when it comes to calls of any kind. Technically flawless and in simple, honest language • Young people want short political videos that are free of technical glitches or flaws, i.e. professional camera work and clear sound. They are not looking for elaborate editing; only 17 per cent want visual filters or effects to be used in the videos. • Of particular importance to them is that the videos contain simple and clear language (68%), that the messages are honest (81%) and that politicians show their human side (62%). November 2025How to Sell Democracy Online (Fast) 88 For young people, TikTok and Instagram are no longer just leisure and entertainment spaces, but also key channels through which they learn about world events. These platforms therefore play a significant role in shaping their (political) attitudes and opinions. Most actors in political representation are now aware of this development. However, moderate-democratic parties in particular still find it noticeably difficult to communicate with young people via short videos on these platforms. The majority of first-time voters say that social media plays a central role in political communication; at the same time, almost as many of them rate the performance of politicians on these platforms as ”poor” or ”very poor” (Institute for Generational Research, 2025). And even though the debate about whether one should be politically active on these platforms at all seems to have died down in 2025, there is still no comprehensive presence of the entire democratic spectrum in this sphere of political communication. Although almost all members of the Bundestag from the democratic parties are represented on Instagram, only 45 per cent are also on TikTok. Members of state parliaments and municipal councils are only sporadically found on the two platforms. Yet democratic engagement usually begins in one‘s own city or county, because decisions made there often directly affect young people. In order to reach them online, targeted progress is needed to secure the reach achieved by moderate-democratic actors during the last federal election campaign and to continue this beyond election campaign periods (Richter et al., 2025). However, since the departure of high-reach ministers from the so-called traffic light coalition government (2021– 2024), such as Robert Habeck, Christian Lindner and Karl Lauterbach, the current federal government in particular has lost significant reach on Instagram, and even more so on TikTok (Netzschreier, 2025) – many incumbent ministers have significantly fewer followers there than their predecessors (Zwick, 2025). Only Chancellor Friedrich Merz stands out, with around 2.6 million followers on Instagram. This is partly because he assumed the followers of his predecessor, Olaf Scholz, with the account; nevertheless, the number has risen since he took office (Zwick, 2025). In order to better understand the expectations of the young target group and how political actors communicate in short video formats, we conducted a study involving three components. Component 1 analyses the spectrum of political short videos on Instagram and TikTok using automated content analysis. In Component 2, adolescents and young adults between the ages of 16 and 27 were asked about their expectations of short videos in several group discussions. Component 3 uses a representative online survey with a selection experiment to record how adolescents and young adults perceive and evaluate political content and which video features on social media ensure that a video is actually viewed by the target group. In combination, these three components make it possible to link structural anomalies in local political communication with the subjective reception preferences and perceptions of the intended target group. The findings of our study are intended to be used to create a cross-party basis for providing young people with democratic information and identification opportunities that meet their expectations. To this end, political actors should be given tools to tailor their messages more closely to the needs and demands of young people – with clear criteria, examples and practical recommendations for action. This should make it easier for young people to access democratic discourse, strengthen their political self-efficacy and thus (re)gain trust in democratic actors. 1. Introduction November 2025How to Sell Democracy Online (Fast) 99 The politicisation of young people For young people, the smartphone is the most important object in their lives (Steinhauer et al., 2025): 96 per cent of 12to 19-year-olds now own their own smartphone, and nine out of ten young people are online every day – on average for over three hours (201 minutes) per day (Feierabend et al., 2024). During this time, they not only maintain contact with friends online, but also use the internet to keep abreast of current world affairs. More and more young people are turning away from traditional media and are instead obtaining their information from social media – where algorithms based on behaviour patterns determine which political content is displayed to users. According to the Reuters Institute Digital News Report 2025, one in two people between the ages of 18 and 24 regularly uses social media to consume news. A third even describe social networks as their main source of information about current world events. Particularly popular news sources in this age group are the short video platforms Instagram, TikTok and Snapchat (Behre et al., 2024; Newman et al., 2025). While, in the 2000s, there was talk of a politically disengaged youth, today we are seeing growing interest among young people in national and global developments. In 2025, 68 per cent of under-30s said they were interested in political issues – an increase of 15 percentage points compared to 2023 (Federal Government Press and Information Office, 2025). However, according to a study by the German Children and Youth Foundation (2025), only 45 per cent of young people believe they can actually influence politics. Only just under a fifth of young people surveyed in a study by the Bertelsmann Foundation believe that they can make a difference on issues that are important to them through political engagement (McDonnell et al., 2024) – and only one in five feels represented by the Bundestag (TUI Foundation, 2025). Feelings of invisibility and powerlessness, as well as a frustrated and alienated perception of a lack of political effectiveness, go hand in hand with young people‘s ambivalent relationship to democracy. Although the majority of young people support democracy (Görtz & Langness, 2024), almost half would like to see a ”strong hand that creates order” (German Children and Youth Foundation, 2025). At the same time, there is a strong desire for participation: young people want to be addressed, heard, involved and taken seriously in politics (McDonnell et al., 2024; Okroi & Wendt, 2024). This must be taken seriously, because political experiences in youth shape democratic engagement throughout life (Kligler-Vilenchik & Literat, 2024). Those who experience positive, sincere approaches to politics at a young age are more likely to be engaged in their adulthood. Conversely, disappointment or political exclusion can lead to long-term withdrawal and disillusionment with politics (Chu et al., 2022; Stattin et al., 2023; Wray-Lake & Syvertsen, 2011). In short: encounters with politics on short video platforms shape the political mindset of the young and thus the next generation. The importance of political communication on social media is illustrative of a comprehensive structural change in the media. 2. Addressing Young People Digitally in Democracy: How to Sell Democracy Online (Fast) – and Why? November 2025How to Sell Democracy Online (Fast) 1616 Specifically, this change means that Meta platforms no longer recommend political posts through their algorithms. On Instagram, for example, this applies to the algorithmically sorted area below the search function, suggestions in the news feed and also to short videos. However, the new rules only apply to accounts that are not actively followed. Anyone who already subscribed to the accounts owned by political actors on Instagram before the platform change in February 2024 will continue to see their posts (Instagram, 2024). Affordances Partly due to its strong focus on visual content, the aesthetics of content are a key affordance of Instagram (Kümpel & Rieger, 2020). The visual logic of Instagram and its perception as a platform, that is supposed to enable a more personal, intimate view of politics (Bossetta, 2023; Haßler et al., 2023), leads to specific communication and mobilisation strategies. Political actors often use the platform to post acclaims, which can be at the expense of political issues (Haßler et al., 2023). They also present themselves as approachable and emotional and grant private insights, but political issues continue to dominate over non-political topics (Achmann-Denkler & Wolff, 2023; Geise et al., 2024). In addition, Instagram tends to favour high-quality content (Chang et al., 2024). Since it is not possible on Instagram to integrate links into Reels, political actors rarely refer to external online sources of information (Wurst et al., 2023). In order to achieve greater visibility beyond their own follower base and to share political content widely, collaborative posts involving multiple accounts are often used (Chang et al., 2024). 3.1.1.2 TikTok Architecture In 2017, ByteDance acquired the Musical.ly platform and launched its successor, TikTok, in the same year. Although Musical.ly was headquartered in China, it had been hugely successful with teenagers in the US and UK since its launch in 2014 and quickly focused its target audience on Generation Z (especially teenagers and pre-school children) (Zeng et al., 2021). TikTok primarily relies on short videos in the form of a vertical full-screen feed that offers users a continuous video stream (Collazos Garcia, 2024). The platform‘s architecture was initially geared towards music and dance, which is why most of the content consisted of imitations and choreographies. Since the merger of Musical.ly and TikTok and the introduction of interaction and editing features, content creators have been able to give their videos context and interpretation. This has also changed the content on the platform, which has become increasingly open to political communication (Hohner et al., 2024). The relatively short duration of the videos, which favours a more informal approach to communication from a political perspective, is said to captivate young users (Medina Serrano et al., 2020). Another key factor in TikTok‘s success is its recommendation algorithm, which promotes a variety of content regardless of followers and accounts followed. The ”For You” page on TikTok algorithmically curates a selection of videos, giving priority to new and so-called trending content over the existing follower base (Hohner et al., 2024; Weimann & Masri, 2023). This is because the TikTok algorithm promotes the virality of content based on engagement metrics rather than follower numbers (Chang et al., 2024). This increases the potential of each video to achieve a wide reach. Through user interactions with videos (including likes, saves, comments, shares) and view time (i.e. how long a video is watched), the algorithmic system collects relevant information about its users‘ preferences and can very quickly present them with ”tailored” content (Schnabel & Berendsen, 2024). This platform logic applied by TikTok encourages content creators to develop videos that incorporate popular features or memes at the time of posting in order to increase their own visibility. November 2025How to Sell Democracy Online (Fast) 1717 Affordances Successful content on TikTok is strongly oriented towards entertainment value: trend sounds, challenges 1 or viral formats are central elements here (Bösch, 2023; TikTok for Business, 2022). TikTok itself explicitly states that creators are more successful when they tailor their content to the needs of their audience, which means that ”simply put, people want their TikTok experience to be fun” (TikTok for Business, 2022). TikTok offers creative tools, such as Green Screen, Duet and Stitch, which allow users to incorporate external elements or respond directly to other people‘s content and process it further. Duet allows users to play their own clip alongside an existing video, and Stitch allows them to embed excerpts from other videos into their own content (Moir, 2023). This specifically encourages interaction with other users‘ content (Quick & Maddox, 2024). Political communication on TikTok thus becomes a far more interactive experience than on YouTube or Instagram (Moir, 2023). Short videos on TikTok are often more spontaneous, informal and cheaper to produce, which allows for a quick response to trends. This contrasts with the more curated and professionalised aesthetic of Instagram (Chang et al., 2024). The architectural and affordance differences between Instagram and TikTok (as well as other platforms) have been little researched scientifically to date. Although interest in political campaigns on social media is growing, there have been few studies on how the design of individual platforms influences the communication strategies of politicians. One reason for this is that social media is often viewed as a uniform medium in research – even though platforms such as Instagram, TikTok and YouTube differ significantly in many ways in terms of their networks, functions, algorithms and how they process user data (Bossetta, 2023). The role of algorithms Although we know little about the algorithms used by Instagram and TikTok due to the lack of transparency on the part of Meta and ByteDance, it is clear that their algorithmic recommendation systems have a significant influence on what content is displayed in young people‘s feeds and how far it reaches (Bösch & Geusen, 2025). Most of the videos that young people see on social networks are suggested to them by the algorithm – so they come across them more by chance than by searching for them specifically (Molem et al., 2024). Two-thirds of young users nowadays consume social media content in this way (Behre et al., 2024). This reliance on algorithms has consequences for how young people absorb political information, because they often encounter news and political topics incidentally and passively in their feed – embedded between entertainment and trends – rather than consciously via the websites of traditional media brands (European Parliament & Ipsos European Public Affairs, 2023). Even though the influence of algorithms can only be measured in detail to a limited degree, findings to date suggest that the recommendation systems of social platforms have a noticeable impact on the visibility of political messages (DeVito, 2017). Populist and radical populist actors in particular benefit from the unknown metrics. An analysis of the 2024 state elections in eastern Germany showed that the AfD dominated the TikTok feed of first-time voters who were less interested in politics (Tjaden et al., 2024). This target group was shown an average of one AfD video per day, while political parties such as the CDU or the BSW only appeared in about one video per week, with other democratic parties achieving virtually no visibility. In some cases, they even posted more content than the AfD. The study suggests that the TikTok algorithm favours certain content – and 1 Challenges describe campaigns (or trends) that invite people to create their own videos by performing a specific task, selecting a specific visual effect, or applying a specific sound (Bösch, 2023). November 2025How to Sell Democracy Online (Fast) 1818 disadvantages the content of moderate-democratic actors (Tjaden et al., 2024). Further studies also show that nationalist and conspiracy-ideological content achieves the highest reach on TikTok (Hohner et al., 2024). The platform algorithm seems to favour content that provokes, emotionalises and polarises (Bär et al., 2024; Boeker & Urman, 2022; Guess et al., 2023; Ibrahim et al., 2025). Populist and misanthropic messages strike a chord, leading to greater engagement on digital platforms and thus a wider reach (Doroshenko & Tu, 2023; Engesser et al., 2017; Jost et al., 2020). However, the academic debate is increasingly discussing ambiguities and warning against false assumptions concerning the effects of algorithms, because many phenomena cannot yet be clearly attributed to recommendation algorithms. Social, cultural and individual factors are closely intertwined with technical factors – and it is difficult to separate their respective causal influences (Dienlin & Johannes, 2020; Metzler & Garcia, 2024; Salganik, 2019). In one case in point, there is debate as to whether social media and its algorithms really reinforce political polarisation – or whether they merely make existing social divisions more visible (Bail, 2022). Algorithm research is still in its infancy, and many studies are limited to individual countries or time periods, rely on self-reported data, and cannot prove clear causality (Molem et al., 2024; Orben & Przybylski, 2019; Sumpter, 2021). In short, current research often cannot determine with certainty whether algorithmic systems cause certain problems or merely reinforce existing social dynamics (Metzler & Garcia, 2024). November 2025How to Sell Democracy Online (Fast) 1919 independent fact-checkers and towards the use of automated filter systems and community notes, which may be formally DSA-compliant but raise serious concerns about transparency and reliability (Cazzamatta, 2025; Trujillo et al., 2025). Critics, such as Reporters Without Borders and the European Federation of Journalists, warn that any such steps could increase the risk of disinformation and jeopardise the role of independent oversight (European Union, 2025). The DSA & TikTok For some time now, TikTok has been deploying an ”Elections Integrity Programme Team” for political elections within the European Union, which randomly checks content to identify risks. In addition, users can hide certain terms in their feed, understand why content is being displayed, reset the algorithm of their feed or manually diversify personalised suggestions (TikTok, 2023). The tech company‘s focus in dealing with elections is almost exclusively on preventing misinformation. However, studies indicate that TikTok‘s recommendation systems can influence political opinion-forming by making content asymmetrically visible and amplifying party-affiliated content differently (Katzy-Reinshagen et al., 2025; Tjaden et al., 2024). Particularly problematic is the fact that TikTok users are less sceptical about possible misinformation than users of other media (Friedrich Naumann Foundation, 2025). This encourages the rapid spread of questionable content – reinforced by the platform design, which allows individual videos to go viral without a follower base, expertise or clear source context (Bösch, 2023). Scientific studies show that, despite official regulations, TikTok does not adequately filter problematic content, such as political advertising or misinformation (Brewster et al., 2022; NYU Tandon, 2022). Legal bases and current debates on political communication on social media The Digital Services Act (DSA) forms the relevant legal basis for political communication on social media. The DSA is a European law that obliges large online platforms, such as TikTok, Instagram (Meta), X and others, to assess the potential negative effects of their algorithms on social debates, electoral processes and public safety on an annual basis (Art. 34 DSA). Particular focus is placed on the risks posed by algorithmic recommendation systems. However, the DSA remains vague about exactly which risks are meant, leaving platforms with considerable leeway in implementing and selecting countermeasures (ISD Germany, 2025; Verwiebe et al., 2024). Although it prescribes the handling of illegal content and the introduction of complaint mechanisms, it does not offer any concrete measures – for example, automated filters or the use of fact checkers are permitted, but their use is not mandatory (Heise, 2024). The EU Commission monitors the enforcement of the DSA for particularly large platforms, while national authorities – in Germany, the Bundesnetzagentur (Federal Network Agency) – are responsible for smaller platforms. The DSA & Instagram (Meta) In the case of Instagram (Meta), regulatory processes focus primarily on the handling of election content and the criteria according to which content is recommended or amplified (European Commission, 2024). Researchers point out that, although Instagram partially complies with the formal requirements of the DSA, there are considerable shortcomings in its practical implementation (Sekwenz et al., 2025). Meta‘s fact-checking programme is also the subject of critical debate: Meta emphasises its flexibility for future adjustments, including in the EU (Meta, 2024), similar to a model newly introduced in the US in January 2025 – away from cooperation with November 2025How to Sell Democracy Online (Fast) 2020 3.2 Research gap: our study The number of studies on political communication on digital platforms has grown steadily over the past few years, reflecting their relevance and use by political actors. While platforms such as Facebook, X and Twitter have been the focus of attention in the past, Instagram and TikTok have gained in importance both in the public sphere and in academia. Nevertheless, studies on Instagram mainly refer to analyses of images (Geise et al., 2024; Haßler et al., 2024), while video formats have not yet been comprehensively examined. In the case of TikTok, public and academic debate in German-speaking countries has largely focused on populist or right-wing extremist actors. Further studies are devoted to reach and interaction metrics, for example, and thus only examine exemplary content (Bösch, 2023; Classen et al., 2024; Meyer et al., 2024; Schnabel & Berendsen, 2024). Little research has been conducted into how moderate-democratic parties address young people as their primary target group in terms of content on platforms such as Instagram or TikTok. As of September 2025, there is only one study in the German-speaking world that systematically examines the content of short political videos from all political parties represented in the Bundestag (Jost et al., 2025). However – like most other studies (Berlin & Fehrensen, 2025; Böhmer et al., 2025; Bösch & Geusen, 2025; Koenigsdorff et al., 2025; Richter et al., 2025; Riedl et al., 2025) – it is limited to the 2025 Bundestag election campaign, while long-term, non-election-related strategies for addressing young people have hardly been studied. The perspective of the recipients has also only been considered selectively so far: there has been little research into how young users perceive and evaluate political content on video platforms and which formats actually have an impact on them (Classen et al., 2024; Tjaden et al., 2024). Although democratic actors are currently playing catch-up with the communication strategy advantage held by populist-radical parties on social media (Berlin & Fehrensen, 2025), there is still a lack of reliable empirical data on which content and formats are successful, whether quantitatively (reach, interaction) or qualitatively (credibility, identification). Moreover, research often limits itself to examining political communication from the perspective of either the sender or the recipient (Tjaden et al., 2024). An integrative analysis that brings both sides together is still lacking. Our study aims to close this research gap. November 2025How to Sell Democracy Online (Fast) 2121 4.1 What politicians send 4.1.1 Political actors examined Accounts owned by politicians and political parties (municipal, state and federal level) We examined all the social media accounts owned by elected representatives at state and federal level from the political parties represented in the last Bundestag: CDU/CSU, SPD, AfD, Alliance90/The Greens, FDP, The Left Party and the BSW. Due to its young electorate and its manifesto focus on the future, we also included the pan-European party, Volt, at federal and state level, as well as its leadership. All German members of the European Parliament were also included. At municipal level, accounts owned by municipal representatives from one independent city and one district were to be selected from each federal state. To this end, a random sample was taken at federal state level to take into account the heterogeneous distribution of districts and independent cities across the federal states. The three city states of Berlin, Bremen and Hamburg were treated as one administrative unit. In the case of Saarland, which has no independent cities, two districts were selected at random. Where the randomly selected representatives of the district or independent city did not have an Instagram or TikTok account, the next post in the random list was selected until an account on Instagram and/or TikTok was found. This resulted in a randomised list of 30 political representatives at municipal level. Three accounts could not be included retrospectively as they were inactive or had been deleted during the study period. We identify three types of political parties that differ in their basic ideological stance on liberal democracy 2 : • Moderate-democratic: The CDU/CSU, SPD, FDP and Alliance 90/The Greens are considered to be firmly anchored in the democratic centre. They pursue a centrist agenda and are clearly committed to the principles of liberal democracy, such as diversity of opinions, separation of powers and protection of minorities (Akkerman et al., 2016; Bale & Kaltwasser, 2021; Decker, 2025; Jolly et al., 2022). Volt can Table 1: Overview of Videos and Accounts by Political Level Level #Accounts Insta #Accounts TikTok #Videos Insta #Videos TikTok Unspecific 117 77 2,668 3,264 Federal 656 285 10,192 4,694 European 67 30 829 322 Municipal 23 3 197 30 State 572 146 6.809 2.134 Total 1,435 541 20,695 10,444 2 The three-stage model is not based on a rigid interpretation of terms, but explicitly recognises fluid transitions between the individual types. For example, even moderate-democratic politicians and political parties sometimes use populist elements in their political address (Decker, 2022; Kaltwasser, 2024) 4. Study November 2025How to Sell Democracy Online (Fast) 2222 also be classified in this category (Knodel, 2023). Due to the political deradicalisation process initiated in recent years, The Left Party is classified as a moderate-democratic borderline case (Decker, 2025; Thomeczek, 2024), as a substantial part of its political ideas are consistent with the above criteria (Meijers & Zaslove, 2021; Rooduijn et al., 2023). With the split of the BSW, a significant pop - ulist faction left the party. We therefore include The Left Party in this group in this study. • Populist: Populist parties present themselves as the sole representatives of the ”true will of the people” against a supposedly corrupt elite. This includes the Sahra Wagenknecht Alliance (BSW), which combines anti-pluralist rhetoric (Mudde, 2004) with a highly simplistic understanding of politics (Decker, 2022; Thomeczek, 2024). • Populist-radical/extremist: The AfD is classified as populist-radical/extremist due to its combination of populist rhetoric and radical and extremist content (e.g. authoritarian and nativist ideology) (Abou-Chadi & Wagner, 2019; Art, 2023; Arzheimer, 2019; Fielitz et al., 2024; Hohner et al., 2024; Rooduijn et al., 2023) – the party is classified as a suspected right-wing extremist organisation by the Federal Office for the Protection of the Constitution (Bundesamt für Verfassungsschutz, 2025). Accounts owned by public authorities This group includes official accounts owned by state institutions at all political levels, such as ministries, offices, state governments and federal authorities. These public administration accounts primarily serve the purposes of information, transparency and crisis communication – they make administrative processes visible (Bornschein, 2020; Haßler, 2024). Accounts owned by party-affiliated youth organisations These include accounts owned by the youth associations of the political parties examined at federal and state level:* • Green Youth (Alliance 90/The Greens) • Jusos (SPD) • Junge Union (CDU/CSU) • Junge Liberale/JuLis (FDP) • Linksjugend [’solid] (The Left Party) • Junge Alternative (AfD) Youth associations are engaged in the civic education, opinion-formation and mobilising of young people, and are often regarded as talent pools for future party careers (German Federal Youth Council, 2021; Gille, 2025). At the same time, they perform a variety of functions within the political parties: they represent the interests of young people, promote political responsibility at an early stage and offer networking opportunities within the party (Bracker & Wibke, 2023; Kaack, 1971). Accounts owned by political influencers Social media influencers are individuals who have built up a large reach through persistent self-marketing on social media platforms (Khamis et al., 2017). On Instagram, they are primarily characterised by high follower numbers, and on TikTok by the reach of their short videos. Originally known from the field of product marketing and entertainment-based communication (Lou & Kim, 2019), today they are increasingly taking on socio-political roles as political influencers without officially belonging to any political party or an organisation (Harff & Schmuck, 2025). The individualised content ranges from commentary on current political events (newsfluencers) to activist or party-political classifications and satirical representations. In a digital media environment that is strongly influenced by algorithmic curation, politi- *The youth association ”Jugendbündnis im BSW” (JSW) was founded in July 2025 and therefore did not yet exist during the period examined (tagesschau.de, 2025b). Volt Violet is a youth network within the party and not an independent youth association under the German party model (European Youth Portal, 2024). November 2025How to Sell Democracy Online (Fast) 2323 cal influencers act as so-called parasocial opinion leaders (Thorson & Wells, 2016). They often convey political content from an everyday, identity-related perspective (Gonzalez et al., 2024; Suuronen et al., 2022) – and, in doing so, bank on an emotional closeness to their followers (Harff & Schmuck, 2025).3 Table 2: Overview of Videos and Accounts by Actor Group and Party Affiliation Party Group of actors #Accounts Insta #Accounts TikTok #Videos Insta #Videos TikTok AfD Youth organisation 10 1 25 2 Party 11 10 219 218 Politicians 87 83 1,383 1,600 BSW Party 9 4 139 137 Politicians 14 5 151 122 CDU/CSU Youth organisation 17 3 85 73 Party 16 8 581 234 Politicians 273 56 3,748 682 FDP Youth organisation 16 5 149 43 Party 16 7 142 39 Politicians 111 33 1,317 496 The Greens Youth organisation 16 8 98 82 Party 17 10 326 265 Politicians 194 84 3,022 1,194 The Left Party Youth organisation 8 6 45 41 Party 15 9 331 141 Politicians 58 26 684 311 SPD Youth organisation 19 5 114 55 Party 17 8 299 264 Politicians 287 102 3,516 1,138 Volt Party 15 4 133 53 Politicians 14 4 100 22 No classification Influencers 40 53 1,694 2,940 Institution 152 7 2,378 292 Politicians 3 - 16 - Gesamt 1,435 541 20,695 10.444 3 Some journalistic – and thus institutionalised – social media formats deliberately imitate the style of political influencers (e.g. the news brand BR24‘s News WG, 181,000 followers). November 2025How to Sell Democracy Online (Fast) 2424 Identifying non-institutionalised political actors is a key challenge in political communication research (Jost et al., 2023). Unlike party politicians or representatives of political institutions, the inclusion criteria cannot be formally defined. This applies both to the distinction drawn between when influencers can be classified as political and to the question of the criteria for including accounts – whether based on reach or taking into account the broadest possible spectrum of opinions. Furthermore, the dynamic developments in the influencer scene make systematic selection difficult, meaning that lists and registers do not provide a reliable basis. The selection of influencers for this study is therefore based on extensive online research. In addition, our project partners, as experts in political communication practice, were asked to review the list and add to it, if necessary. Although it is not possible to assign influencers directly to political parties, the phenomenon of ”TikTok guerrilla” (see Chapter 3) was taken into account: around 20 per cent of influencers can be classified as positioning themselves close to the AfD. Even though we have taken the utmost care and incorporated relevant perspectives in our efforts to make a well-founded selection of influencers, this selection does not claim to be representative. When interpreting the results, it is important to bear in mind that the communication of ”the” influencers might have been different had a different selection been made. A complete list of the accounts examined can be found in Appendix A1 for classification of the findings. 4.1.2 Content analysis Our study is based on an AI-assisted content analysis of short videos from political actors and influencers on Instagram and TikTok, whereby we randomly selected every other post from all videos posted between June and December 2024 (Ntotal= 31,139). The categories surveyed concerned topics, youth relevance, tone of voice (acclaims; attacks on political opponents), direct address, calls to action and platform-specific affordances. To this end, a codebook was first developed, on the basis of which three trained coders manually analysed between 200 and 1,000 videos per category. Intercoder reliability was above Krippendorff‘s alpha of .67 in all categories. This data was then used as the gold standard for the subsequent automated content analysis. The actual analysis was performed using the Gemini 2.5 Pro language model, which supports multimodal inputs, thus enabling the combined analysis of visual and textual elements. To this end, the codebook created from the manual coding was translated into prompts for the language model – both in the form of classic instruction prompts and structured ”tree-of-thought” (ToT) prompts, based on approaches by Hulbert (2023). The prompts included both the video and the corresponding caption; in some cases, few-shot examples were added to improve classification accuracy. The model outputs were systematically compared with the results of manual coding, with particular focus being placed on rare categories. Deviations led to an iterative refinement of the prompt strategy, which was then validated again on a manually coded training dataset, achieving largely satisfactory F-scores. Categories that were rarely present in the coded material and that are predominantly identified via movement sequences proved particularly problematic. When categorising the topics, we followed the classification of the topic areas used in the German Policy Agenda project (Breunig et al., 2024), aggregating the topics into their supercategories (F1 = .70). The category ”youth reference” was used to record whether the video made explicit reference to young people, chil - dren, adolescents, young adults, future generations and their perspectives, concerns and needs and/or their future, or whether this reference was made by directly addressing or naming the target group (F = .78). Attacks on political opponents (negative campaigning) were coded when negative statements referred to political parties or politicians, including attacks on their traits, positions, past or future achievements (F = .78). Acclaims were coded when positive statements referred to political actors, including positive statements made by political parties or politicians November 2025How to Sell Democracy Online (Fast) 2525 in terms of their traits, positions on political issues, past or future achievements (F = .66). Moreover, any time that the speaker or creator of the video directly addressed recipients through speech or text (F = .87), this was recorded. Whenever a form of direct address and a specific call to action occurred, the type of calls to action was coded. The call to vote (F = .88), calls to join a party (F = .75), calls for donations (F = .80), as well as other calls for offline actions (e.g. participation in demonstrations; F = .76) and online actions (e.g. sharing content, commenting; F = .76) were recorded. Platform-specific affordances, such as dancing (or movements to sounds and music; F = .44), interaction with text elements (e.g. pointing to text elements; F = .40), as well as the display of infographics (F = .56), the use of a split screen (duet; F = .50) and the recording mode (selfie, F = .68), were also recorded.4 4.1.2.1 Content In 2024, political actors were very preoccupied with themselves in their communications Political communication in short videos begins with a simple but crucial question: What is the subject matter? Hence, before we evaluate how content is staged or what strategies are used to convey it, we first analysed the topics found in the TikToks and Reels posted by political parties, politicians, youth organisations and influencers. The analysis is based on the assumption that certain topics are particularly relevant to young people, either because they are directly affected (e.g. education policy) or because of the significance of the topic for their future (e.g. environmental and climate policy). One in three TikTok videos or Instagram Reels from politicians revolves around the actions of the government or other political institutions, or is dedicated to topics (at the time of data collection) related to the upcoming 2025 federal elections (Fig. 1). Videos concerning current political issues, such as law and crime (just over 6%), migration (just under 6%) and the social welfare state (just over 5%), rank behind videos with no clear political reference, at around 8%. Topics that are particularly relevant to young people – such as citizens’ rights, work or housing – play a significantly smaller role in the overall picture and all remain in the range of around three per cent of posts. Apart from the stronger focus on election topics on TikTok, only minor differences were identified between the TikTok and Instagram platforms. A look at the topics addressed by the political parties (Fig. 2) shows that, in addition to government actions and mistakes, the political parties, in their political discourse, especially highlight those policy areas in which, according to issue ownership theory, the population already attributes particular competence to them (Petrocik, 1996). As such, the AfD communicates the two top issues (government & administration and elections), followed by a high proportion of videos dealing with migration (17%). In doing so, it focuses much more strongly on law and crime than the other political parties. The BSW stands out for its high proportion of videos on defence (15%) and international politics (14%). The CDU/CSU has run with twice as many non-political videos (12%) as it has videos on economic issues (6%), whilst the FDP focuses particularly on the economy (13%) in addition to government and administration. At 10 per cent, the environment plays a more central role for The Greens than for any other party. In addition to the two dominant topics, The Left Party places a clear focus on the social welfare state (14%) as well as housing (7%), whereas the SPD stands alone in exclusively highlighting the social welfare state (10%). Volt is an exception in terms of party communication, 4 To determine the F-scores, the manually coded data was divided into a test set and a training set. This has a negative impact on the F-scores, especially for categories that occur infrequently. When all data was taken into account, the values for the categories ”dancing” and ”infographics” improved (F(dancing) =.67; F(infographics) =.61). However, the reliability for the categories ”interaction with elements” and ”duets” remains problematic. Against this backdrop, any interpretation of the corresponding evaluation should be viewed with caution. November 2025How to Sell Democracy Online (Fast) 3232 above all, for their high rate of online calls to action (44%), while calls to vote play virtually no role here, at 1 per cent. Politicians remain cautious across all categories, reaching only 8 per cent in calls to vote. Influencers are the only group that advocates financial contributions to any significant extent (4%). The party-specific differences extend to the calls to action (Fig. 7). Volt leads in all categories: they call for offline and online actions particularly often (39% each), rely on calls to vote more often than average (28%), advocate party membership most frequently (8%) and ask for donations the most (2%). Although The Greens rely less often on calls to vote than Volt (around a fifth of their posts), they still rank ahead of the CDU/CSU and the BSW and are in the upper midrange for offline and online actions. The CDU/CSU has a rather cautious mobilisation strategy: they have low figures for calls to vote (6%) and online actions, but slightly higher figures for offline actions. The AfD uses calls to vote slightly more frequently than the CDU/CSU, but lags behind in calls for donations. The FDP has a distinctive feature: it completely refrains from any calls for donations and is also poorly represented in terms of calls for party membership and offline campaigns. Overall, the BSW mobilises the least – both direct address and all other forms of call to action rank well below average. In general, calls for offline action tend to be more prevalent in the centre-left of the political spectrum. Calls of this kind accounted for over 20 per cent by Volt, The Left Party and The Greens. Barely any dancing: generally little creativity Senders of short videos can build a relationship with consumers of video content not only through direct addresses or calls to action, but also through so-called affordances. These serve to facilitate interaction with and between users, but also content creation itself (for more information, see Chapter 4.1.1) and are often platform-specific, even though Instagram and TikTok are now similar in many ways in terms of their design options (see Chapters 4.1.1.1 and 4.1.1.2). November 2025How to Sell Democracy Online (Fast) 3333 The political parties, youth organisations, political institutions and influencers we analysed most frequently use the selfie format, which appears in 17 per cent of all videos. On TikTok, it is used more than twice as often as on Instagram (27% compared to 12%). Infographics appear on both platforms in only 7 per cent of all videos. Interactions comprising elements – such as pointing to text or graphics that appear on screen – are found in only 4 per cent of all posts. On TikTok (7%), this occurs more than twice as frequently as on Instagram (3%) (Fig. 8). Contrary to popular belief, dancing is rather rare. Dancing or other choreographed movements appear in only 2 per cent of all videos, with TikTok (2 per cent) only slightly ahead of Instagram (1 per cent). In terms of dancing, youth organisations (6%) and influencers (4%) rank ahead of the other groups, which hardly ever use such formats (1–2%). The duet feature, which allows users to combine an existing video with their own video side by side, is used in one per cent of videos overall, with slightly higher usage on TikTok (2%) compared to Instagram (1%). Comment stickers are also rarely used (1% of all videos), with TikTok (2%) slightly edging out Instagram (1%) here. Unsurprisingly, the group whose main work consists of creating short videos makes the most use of technical and creative possibilities: political influencers. They employ the selfie format by far the most, in just under half (47%) of their posts. This form of self-expression is not only the highest among all affordances, but also a clear outlier compared to other groups – youth organisations record 16 per cent of their videos in selfie mode, politicians 13 per cent, political parties 9 per cent and institutions only 3 per cent. Interactions using visual elements, such as text and images, are also particularly typical for influencers (13%), while they occur much less frequently in videos from all other groups (3–5%). Deviations in the use November 2025How to Sell Democracy Online (Fast) 3434 of infographics were ascertained: institutions use them most frequently (12%), followed by influencers (8%), political parties (7%) and youth organisations (7%). Politicians employ them least frequently (6%). Political parties are much more cautious in their use of the platforms‘ design options. Looking at the differences between them (Fig. 9), the following stands out: The Greens, The Left Party and, in particular, Volt make the most use of infographics (9%, 9% and 11% respectively), while the CDU/CSU and FDP use this format much less frequently (4% each). Volt also stands out well above the others in terms of the selfie format: around a third (32%) of its posts use this personalised form of presentation, followed by The Greens (18%) and the AfD (14%). The Left Party (8%) and the BSW (9%) are the least likely to record videos in selfie mode. Interactions using visual elements, such as pointing to superimposed text, are mainly found among The Greens (5%), while the CDU/ CSU and AfD rarely use them (2% each). Dance or music-based movements are rare overall; only Volt (2%) and The Left Party (also 2%) achieve slightly higher values here. Overall, political actors make little use of the technical and design possibilities offered by the platforms. Volt and The Greens show the greatest affinity for this, while the CDU/CSU, the AfD and the BSW are the most reserved in terms of visuals and interactivity. Selfie mode, which is still the most frequently used feature, can be both a conscious means of creating a sense of closeness and an indication of a tight budget for content creation or limited time resources for content editing. 4.1.3 Success factors The analysis of around 31,000 short political videos from six months provides insight into which content November 2025How to Sell Democracy Online (Fast) 3535 and formal characteristics influence the reach of videos, as measured by their number of views. In short: Which topics and which style bring more or less visibility? To this end, a negative binomial regression involving varying intercepts was performed for the individual profiles (Fig. 10). This allows us to estimate the effects of the content independently of the characteristics of the profiles (e.g. number of followers or party affiliation). Attacks bring reach, and nobody likes a bragger The use of attacks (ranging from criticism to insults) had the strongest positive effect on the average number of views: videos attacking political opponents are viewed on average around 40 per cent more often than comparable content without such attacks. It is important to note that views are not synonymous with approval and that the subject target group of our study, young people, tended to reject the attacks in the videos during the selection experiment (see Chapter 5.2.1). In contrast, acclaims significantly reduce reach: videos in which political parties or candidates praise themselves or their allies receive, on average, about 12 per cent fewer views than comparable videos not employing this strategy. Social issues are unpopular, migration is the top issue Among the main topics, that of migration achieved the highest gains: videos focusing on this issue were viewed around 11 per cent more often on average. The attitude adopted by those behind the video was not recorded. Posts concerning elections (8 percentage points more) and videos without any political reference (9 percentage points more) also significantly increased reach. In turn, focusing on social policy (7 percentage points less), the environment (18 percentage points less) and education (17 percentage points less) has a detrimental effect. In addition to algorithmic influences, topicality also plays a role here. Even though these topics, which are often considered relevant to the young target November 2025How to Sell Democracy Online (Fast) 3636 Note: Figure showing regression coefficients and confidence intervals; red estimators indicate a negative correlation between content characteristics and engagement, green estimators indicate a positive correlation, and grey estimators indicate no significant correlation. group (McDonnell et al., 2024; Schnetzer et al., 2024; Weber, 2024), receive fewer views on average, explicitly referring to the concerns and perspectives of young people increases the reach of short videos by around 5 per cent overall. By contrast, the use of infographics has a negative effect on their reach, while other interactive elements have no significant impact on reach. Overall, the results suggest that a confrontational style and the inclusion of the audience in dialogue – for example, through direct address and calls to action – contribute more to reach than factual, informative formats or conventional acclaims. It is not clear why they are viewed more frequently. In all likelihood, the algorithmic settings of the platforms, which favour the visibility of certain content, play a key role. The behaviour and interest of (young) recipients also impact reach. 4.2 What young people (don‘t) want The videos considered successful on Instagram and TikTok because they have a lot of likes and views do not automatically mean that young people want to watch them. Many factors, such as the algorithm, users of other age groups and monetary advertising, can also have an impact on the reach of a video. The black box is large and contains many unknowns. In order to make an informed statement about which videos are actually successful due to algorithms and which are successful because they are tailored to the target group, it is important to engage in dialogue with the target group itself. It is therefore essential to ask young people what they expect from political content in short video format and what political content they do (not) want to see on the platforms. November 2025How to Sell Democracy Online (Fast) 3737 With this in mind, we conducted four focus groups comprising a total of 27 young people in the age groups 16 to 19 and 20 to 25. The concept was developed and implemented in cooperation with Pollytix Research. In order to obtain representative responses about the ideas and expectations of young people in general, we supplemented the qualitative survey with a representative interview. To this end, we recruited a sample of 1,748 young people aged 16 to 27 (M = 22.64; SD = 2.95), representative of men and women in this age group, through the panel provider Bilendi. Approximately half of the respondents were male (52%), 48 per cent female and 0.8 per cent diverse. Around one third (36%) of all respondents had a migration background. Just under half a high level of education (49%), 35 per cent a medium level and 16 per cent a low level. The political orientation of the young respondents essentially corresponds to voter distribution within this age group in the last federal election in 2025: relatively broad, with a clear focus on The Left Party (24%) and the AfD (16%) (tagesschau.de, 2025a). Less strongly represented is the inclination to vote for the CDU/CSU (13%) and SPD (12%), as well as Alliance 90/The Greens, which is preferred by 9 per cent of respondents. One in ten states that they have no affiliation to any party. A similar percentage of respondents indicates that they did not vote in the last federal election despite being eligible to vote (9%). According to the respondents, slightly more than three-quarters (76%) voted, which means that voting behaviour among the respondents is close to the last turnout at the last federal election (Demografieportal, 2025). Self-assessment: interested in politics, but not heard As in other recent studies, the general interest in politics among the respondents is rather high: around half (50%) show strong or very strong interest, while only 4 per cent say they have no interest at all. The interest of young people is clearly highest at national level (64% strong or very strong), followed by international politics (53%). Interest is significantly lower at regional (42%) and local level (37%). In general, young people‘s self-assessment of their political awareness is predominantly high: 65 per cent rate their knowledge as good or very good and 25 per cent as less good. Only around 5 per cent rate their own political knowledge as poor and just under 5 per cent are unable to assess it. However, their perception of their own level of awareness does not necessarily mean that their information base is actually comprehensive or balanced. For example, those who consume political news primarily via social media cannot automatically expect balance, source transparency and contextualisation. When it comes to political self-efficacy (Fig. 11), there is a clear difference between the perception of one‘s own responsibilities and capacities (internal political self-efficacy) and the assessment of one‘s perceived potential influence (external political self-efficacy) (Beierlein et al., 2012; Campbell et al., 1971; Vetter, 1997): While 51 per cent state that they understand and can assess important political issues well, and 46 per cent feel confident about participating in political discussions, only 18 per cent see a close connection between politicians and the population. Similarly, few (17%) agree somewhat or completely that politicians care about the concerns of ”ordinary people”. In our focus groups, young people also express the following feelings of low external political self-efficacy: ”So, yeah, I do [feel politically addressed]. But, then, these aren’t politicians in the Bundestag, ‘cause politicians that might appeal to young people don‘t get into the Bundestag.” (male, 23, student) ”I’d just say that, since many political parties tend to focus more on older generations [...], we younger people, like, we don‘t really have a voice in politics. So, that [sic], there are very few political parties or party manifestos around that appeal to young people.” (female, 18, school pupil) November 2025How to Sell Democracy Online (Fast) 3838 Question text: ”Now we would like to ask for your opinion on political issues. You may agree more or less with the following statements. To what extent do you agree with each statement?” ”I think they maybe just don‘t care about the issues that interest us [young people].” (female, 24, employee) The feeling of not being heard or involved not only leads to low political self-efficacy among young people, but can also cause frustration and a loss of trust. This frustration is not limited to political communication, but also extends to the perception of political outcomes and processes: ”It‘s disenchanting, ‘cause you get the feeling that somehow nothing really changes. So, only they say it though, but what happens next is something completely different” (male, 21, apprentice) ”Sorry, but I just don‘t think anything gets done. [...] So, for me, absolutely nothing’s changed as I see it. So, rather [...] negative, ‘cause people [are becoming] more and more dissatisfied and nothing’s changing. I just don‘t see it.” (female, 17, school pupil) Social media is the most important source of political information – ahead of family, friends and school Young people – regardless of their educational background – say they most often come into contact with politics online (74%) (Fig. 12). This is ahead of family (58%), friends and acquaintances (54%) and educational institutions (60%). At an average of 46 per cent, traditional offline media lag significantly behind the use of digital offerings. Young people use social media platforms almost as often as messenger services (WhatsApp or Telegram, at least daily: 84%) when they are online (Fig. 13). Instagram (81%) and TikTok (61%) top the list, followed by YouTube (58%) and Snapchat (48%). Facebook (19%), Discord (15%), X (15%), Twitch (11%) and BeReal (9%) are used significantly less frequently on a daily basis. Social media therefore plays a decisive role in young people‘s political awareness. They recognise this themselves: ”So, if I want to find out about [...] political parties, for example, then, to be honest [...], I‘m just too lazy to somehow or other read through the election manifesto and [...] then I’ve actually gone on TikTok and, like, had a look at a brief description about what they want, you know.” (female, 17, school pupil) ”I have to admit, I‘m a [...] brutal social media victim. So, when I get wind of something, it‘s really only through social media platforms like Instagram or Twitter or whatever.” (male, 23, unemployed) November 2025How to Sell Democracy Online (Fast) 3939 Three in five young people follow political content – via influencers Comparatively few of our respondents specifically follow politicians‘ accounts on social media (38%), while significantly more young people follow political influencers (60%). Better educated young people (41%) and respondents aged 20 and above (39%) follow traditional politicians slightly more often than other groups. At the same time, they are less interested in political influencers: 56 per cent of those with a higher level of education and 58 per cent of older respondents say they do not follow them. The difference to 16to 19-year-olds is particularly striking – around two-thirds (68%) of them follow political influencers. The focus groups make clear that young people often see them as ”more independent” voices, some of whom enjoy more trust than politicians. Young people accuse politi - cians of using their accounts to cultivate their own image and stage themselves. To counteract this, two common strategies appear to be adopted by the young respondents in the focus groups: either they do not follow any official political accounts at all, hoping that this will prevent them from being influenced in their opinion-forming, or they consciously follow a broad spectrum of politicians from different political parties in order to escape what they perceive as filter bubbles: ”So, I actually try to follow as many as possible, like, both The Left Party and the AfD, so I don’t get caught up in this filter bubble. But it’s hard to completely avoid it really. But I actually follow quite a lot. But, like, not just one political party or specific politicians, but actually all sorts. Even if it doesn’t match my political views” (male, 17, student). ”[...] I don‘t actually follow politicians or any influencers, ‘cause, in some way, I think it might influence my opinion a little bit.” (female, 19, apprentice) Question text: ”Where do you encounter political issues in your everyday life? Please select all areas that apply to you.” November 2025How to Sell Democracy Online (Fast) 4040 Question text: ”You stated that you also use social media. How often do you use the following platforms? I use...” ”So, if I was looking for anything like that [political content], I’d want to get it from influencers or people like that, but not directly from political parties. ‘Cause, like we discussed [...] earlier, you can‘t trust them and you can‘t take them seriously, and all that. ‘Cause you just assume anyways that they’re just doing it to win over voters and won‘t stick to anything they said once the election is over. That‘s why, if there’s people I don‘t trust, I don‘t really want to hear from them. Then, I’d just somehow want to hear what objective people say and think that are not linked to these, these political parties” (male, 23, student) ”For me, the only way I can somehow engage with politics [...] is through social media. I don‘t think you should trust everything on there, because, like I said, everything’s only ever shown from its best side.” (female, 24, employee) What appears in the feed is viewed Young users of short video platforms often watch what is played to them via their personal feed instead of specifically searching for actors and their posts. Half of those surveyed say they often or very often stumble across political content that grabs their attention (47%). As a rule, they then watch this algorithmically selected content. The main problem here is the so-called ”news-finds-me” effect: many users believe that they no longer need to actively search for news because important information appears in their feed anyway. However, this reduces their interest in November 2025How to Sell Democracy Online (Fast) 4141 deeper political knowledge and, at the same time, users rely on a passive, random or algorithm-driven supply of information (Gil De Zúñiga et al., 2017; Song et al., 2020). As such, it makes little difference whether political actors or influencers are posting. Young people are well aware of the intentional and unintentional presence of political content in their social media feeds and evaluate it differently: ””I only follow friends, ‘cause I know I’m going to get people with political content shown to me anyway, ‘cause I leave likes on their videos. I rely a bit on my algorithm ...” (female, 17, school pupil) ”It‘s really convenient that you just get to see this stuff in passing without having to actively search for it” (female, 20, student) ”So, when I’m scrolling through TikTok or something, I’ll get one or two political TikToks. I actually see that as a problem, too. Because TikTok has an algorithm that shows you what you like. And if, like, I say I like left-wing parties, the algorithm recognises that and only shows me those political parties or those opinions and nobody else’s.” (male, 17, school pupil) However, only a minority of young viewers actually interact with the posts that are algorithmically curated and displayed in their feed (Fig. 14 & Fig. 15). Less than a quarter say they like or comment on the posts (influencers 24%; politicians 21%), and only just under one in six participates in an online discussion (17%) or shares these posts on their own profile (influencers 17%; politicians 15%). Just under half of those surveyed (influencers 44%; politicians 48%) never do so. There are virtually no differences between videos from political influencers and political actors. Wanted: short political videos on social media platforms Just under two-thirds of respondents (64%) clearly and unreservedly see social media as a suitable space for reaching young people politically (Fig. 16). Only 13 per cent disagree with this statement and only 18 per cent think that political content is out of place on social media because these platforms are a space exclusively for young people. A majority (56%) believe that political contact on social media Question text: ”Think about the social network you use most often as a source of information. To what extent do the following statements apply to you or not?” November 2025How to Sell Democracy Online (Fast) 4848 target audience. This is because addressing the audience directly (e.g. ”To you young people ...”) increases the chances that the video will be watched further by 17 per cent. This preference for personal references is already evident in the focus groups: ”Then I’d like there to be an example or something at the beginning that draws you in, you know?” (female, 23, student) ””A one-on-one chat, sort of. [...] An honest conversation, I’d say.” (male, 23, unemployed) Interestingly, however, any reference to young people or to the concerns and interests of future generations has no significant influence on the address. Young people want to be addressed, but not in a way that is specific to their age group or with promises for the future. Please don‘t dance, thank you Neither calls to vote nor appeals to join a party or take action online or offline have a significant impact on the probability of selection. This may also be due to the fact that the experiment was conducted after the 2025 general election and interest in and the perceived relevance of calls to vote may have waned. Of the typical TikTok gestures and actions tested, only dancing or movement sequences have a significant effect – albeit a clearly negative one: if dancing is shown in videos, the probability of selection decreases by 45 per cent. This result indicates that young people really want political actors to deliver political content via short videos. Videos that deal with the topic of migration have a greater chance of being selected. On the other hand, videos without a clear political theme or with no reference to the topic of elections are unappealing to respondents. Reasons for selection: topic and storyline appear to be the focus When it comes to selection decisions, clear parallels arise between the reasons for and against continuing to watch short videos. Respondents who decide to watch a video most often say that the video appealed to them (50%). However, it remains unclear what exactly appealed to them. The equally high level of interest in the topic (46%) and curiosity about how the story unfolds (39%) indicate that content and suspense play a decisive role. Inversely, respondents opting not to watch any of the proposed videos most frequently state that the videos did not appeal to them Question text: ”Imagine you saw this video on Instagram or TikTok. How likely are you to do the following after watching this post? Select all options that apply to you. I would...” November 2025How to Sell Democracy Online (Fast) 4949 (74%). This may be for very subjective reasons, but here, too, a lack of interest in how the story continues (30%) and the topic (28%) are key factors, once again highlighting the content as the relevant point of reference. Other, less frequently cited reasons on both sides include liking or not liking the people shown (25% vs. 19%) or a positive or negative perception of the editing (22% vs. 15%). The prominence or lack thereof of the individual (17% vs. 13%) and emotional reactions (15% vs. 11%) play only a marginal role in both directions. November 2025How to Sell Democracy Online (Fast) 5050 The following recommendations for action are intended to provide moderate-democratic actors with evidence-based strategic suggestions and practical tips for increasing their visibility on short video platforms and aligning their communication more empathetically with the assessments and needs of young people. Young people should be addressed where the majority of their reality and sources of information lie – primarily on TikTok and Instagram – at all levels of political framing: federal, state and municipal. Local politics in particular often affects young people directly; for many, involvement in youth parliaments or state youth organisations is a gateway to democratic action. The recommendations do not promise guaranteed virality; platform dynamics remain complex and all too often inscrutable. They are based primarily on the available findings and aim to systematically identify and utilise effective levers rather than following shortterm trend logic. The strategic task is to respond sensitively to the needs and wishes of younger generations on social media and thereby develop effectiveness. In practical terms, it is important to gain a better understanding of the architecture and affordances of the platforms and optimise them in terms of organisation, technology and design. 1. Political presence on TikTok and Instagram: not an ‘if’, but a must For many young people, short video platforms are a natural and prioritised source of political information and opinion-forming. Political actors therefore need to maintain a continuous presence on these platforms – at federal, state and municipal level. At the same time, critical monitoring of the platforms (regulation, law enforcement, standards for the protection of values and young people) should not slacken; democratic meeting places must also comply with legal norms and social values there. However, presence alone is not enough, because it is the how that matters. Differences in platform target groups, architecture and affordances should be taken into account. Essentially, it is about successfully conveying politics to young people in the social media space. Every communication needs a strategy. Before launching an account, goals, target groups and key topics must be defined, ideally in close dialogue with the target group. Processes for channel moderation, labelling and legal notices must be clarified in advance. 2. Authenticity is the most important thing Authenticity arises through consistency, intimacy, everydayness and immediacy, and acts as an anchor of trust.7 It is strategically more important than trying to imitate an artificially ”young” style, 6 The practical tips were developed in cooperation with communication experts accompanying the study as well as on the evidence-based wishes of young people. 7 According to communication scientist, Simon Lübke (2022), authenticity in politics is demonstrated by whether someone pursues a clear line across different appearances (consistency), allows personal insights (intimacy), appears down-to-earth and grounded (everydayness) and also appears spontaneous or unpolished at times (immediacy). 5. Recommendations for Action and Outlook Strategic notes Practical tips6 November 2025How to Sell Democracy Online (Fast) 5151 because the young target group notices and disapproves when a person pretends to be someone they are not. External feedback (preferably from young people) is useful because self-assessments of one‘s ”own authentic style” can easily be misleading. It is advisable to choose an appearance that suits the person and only to employ creative elements (stickers, music, duets/stitches) if they support the content and personality. Young people rarely fail to notice when someone shies away from using the platforms and short video format. Inauthentic appearances, such as stiffness or artificiality, can cause messages to be avoided, misunderstood or even interpreted as an inappropriate intrusion into the reality of young people‘s lives. If personalisation does not work, topic-specific clips or employees who are comfortable with the format can be deployed as an alternative. Short videos that successfully appeal to young people do not have to contain youth slang or include cutscenes with memes. Even serious or ”old-fashioned” camera appearances, recorded in one take, can be just as successful as modern and humorous posts, as long as they are credible and interesting to the target audience. In this respect, it is worth taking the time to evaluate which format, style and content genuinely suits the protagonist owning the account and have recognition value. It is advisable to create a short style guide with examples of do‘s and don‘ts and a recurring format for recognition, as well as to establish feedback loops involving young test groups. Example: Former Vice-Chancellor, Robert Habeck (Alliance 90/The Greens), usually speaks objectively and directly into the camera on his accounts, using only a few design elements. His stylistically quite simple statement video, which he posted three days after the 2025 federal election, achieved more than 600,000 views. 3. If you want to be understood, speak clearly Comprehensibility is a question of democratic access. Complex goals and plans require plain and simple language and a structure that addresses self-efficacy without veering into populism. It is important to clearly demonstrate what relevance the topic addressed has for the everyday life of the target group. Abstract plans for the future can seem all too distant. Those who do not feel addressed and engaged will switch off – or swipe on Technical terms, foreign words and long sentences are only helpful if they are explained precisely. The structure of the video is also crucial for comprehensibility. Content should follow a clear sequence (e.g. problem – goal – solution or oddity – fact – resolution – explanation). Structural editing, short overlays and consistent use of subtitles further improve the structure. The introduction can be planned as an emotional (but not populist) hook (e.g. referencing realisation, participation or control). Image-sound coherence and short explanations of technical terms are equally important. Subtitles not only have an inclusive value, but also a practical one: they make it possible to watch videos in noisy or public environments. 4. Avoid clumsy acclaims; link successes to content Past achievements are particularly relevant in election campaigns, but they are only credible if linked to specific content and real-life situations. A down-to-earth or emotional presentation of achievements that is well-founded and factual can generate interest and strengthen trust, whereas exuberant, triumphalist communication can leave the impression that politics is all about winning and losing. Successes should be presented as short cases (e.g. following the sequence of initial situation – activity – impact on everyday life). Superlatives should be avoided and evidence should always be November 2025How to Sell Democracy Online (Fast) 5252 provided. Self-reference should be limited, while the focus should be on information and the benefits for young target groups. Finally, the individual(s) responsible for the success must be clearly communicated. 5. Taking young people seriously means taking their reality seriously There is no such thing as ”young people” or ”youth issues”. Relevance arises when posts are linked to current living conditions and concrete political actions; distant promises for the future alone often seem remote and intangible. Topics should be planned around typical life situations, such as school, training, studying, starting a career, housing or mobility. Explicit references to young and future generations should be used situationally, and the immediate relevance should always be highlighted, for example: ‚What are we doing today to solve X?‘. Videos with explicit references to young or future generations usually achieve greater reach. 6. Social issues need special care Social, welfare, environmental or health issues should not be neglected due to any focus being placed on polarising, hype-driven topics that dominate the agenda of the time. Credible voices and links to everyday life increase visibility without resorting to sensationalism. It is therefore important to consider how and by whom these topics are discussed. For greater visibility, it is advisable to involve young elected representatives, young people or youth committees as co-narrators. Demands, progress and grievances should be presented concisely and verifiably, with clear references to the everyday lives of young people in the here and now. It is also advisable to cooperate with suitable influencers if they can open up access to relevant target groups and draw attention to the issues in a creative way. 7. Criticism yes, but with style and substance Despite algorithmic negativity bias, young target groups expect a factual, professional tone. Escalation is legitimate if it is comprehensibly justified; personal attacks are counterproductive. It makes sense to classify third-party statements step by step using stitches or duets, to structure counterarguments and to cite sources in order to spark debate. It is also advisable not to con - fuse entertainment value with political appreciation, but to make deliberative standards visible. This not only sharpens one‘s own position, but also supports the formation of differentiated opinions; at the same time, this approach comes across as respectful and credible to young target groups.. 8. Countering radical populism and extremism Populist or extreme videos often achieve a wide reach on Instagram and TikTok. However, this is not necessarily because the young target group likes them. Their comment sections are often used for heated discussions; these reactions and interactions ensure that the video is favoured by the algorithm. According to their own statements, young people value honesty, reject clumsy acclaims by politicians and do not want others to be disparaged. Moderate-democratic actors can use this knowledge and calmly point out how extremist actors are dishonest, self-promoting and aggressive. It is particularly effective to plan responses in concise formats, such as fact checks or ‚myth/ reality‘ comparisons. Personalised attacks should be avoided; instead, it is advisable to address positions and evidence directly. Stitches and duets offer good opportunities to directly address and refute problematic content. They can also transform a simple post into a discursive commentary, thus contributing to the formation of opinions based on different perspectives. It is also advisable to define the frequency November 2025How to Sell Democracy Online (Fast) 5353 and criteria for responses in advance and to link to further resources. Example: For 100 days leading up to the European elections, Member of the Berlin Parliament, Marcel Hopp (SPD), daily refuted one statement made by the AfD (e.g. that there is no gender pay gap between men and women in Germany) in a short video on TikTok and Instagram, in which he displayed written or verbatim quotes from the political party, critically assessed them and countered them with facts. 9. Selfies, filters, speed: what counts is the beginning and credibility Recognisability and basic quality are more important for building a social media account than creative abundance. The survey and experiment show that, as a rule, less is more. Political actors are not expected to fully exploit and utilise the technical and creative possibilities of the platforms. Filters can even contribute to scepticism, while selfie mode can convey authenticity. However, the same applies here: authenticity wins out. The first few seconds also determine whether a relevant part of the target group continues to watch (Anggraeni & Rofiah, 2024; Lee et al., 2025). A short, personalised intro can serve as a recognition anchor. A clear hook should be set in the first three seconds. Equally important are clear sound and a steady image; when editing, it is advisable to stick to one message per sequence. A consistent concept makes it easier for users to find their way around. Example: Foreign Minister, Johann Wadephul, uses recurring formats and categories, such as Q&As, country briefings and the section ”Is Russian disinformation still so crazy?”. 10. Democratic interactions: mostly theoretical so far Unlike influencer communities, political videos do not thrive on intensive interaction. It is striking that young people appreciate the opportunity for direct contact via social media in theory, but rarely use it in practice. Political actors could certainly try to understand the reasons for this and focus on more intensive exchange so that young people feel more noticed and are not afraid to get involved in politics. This applies in particular to issues that directly affect young people and where the impression is created that they are merely being talked about. A current example is the debate on military service or the ban on smartphones in schools. To encourage interaction, it can be helpful to offer time-limited Q&As or digital consultation hours, conduct surveys to prioritise topics, and provide a clear feedback channel, such as a contact form or office address. It is also important to be transparent about the results and progress of surveys. Questions can be submitted anonymously, read aloud and then answered for everyone. In terms of tone of voice, it is advisable to use a conversational rather than a lecture-like style: dialogical, calm and appreciative. 11. Influencers – access via trusted individuals Many young people tend to follow political influencers rather than political parties, as they perceive them to be more independent and honest. The creative formats of influencers‘ videos are also generally rated more positively than those of politicians. In short, targeted collaborations can provide access to new groups, provided that the target group is tailored, values are aligned and transparency is guaranteed; even without formal cooperation, open communication of information is worthwhile. It makes sense to establish clear cooperation criteria and transparent labelling practices. November 2025How to Sell Democracy Online (Fast) 5454 Suitable partners should be carefully vetted and key messages should be developed jointly. Influencers and creators should be given considerable freedom in their design within the framework of defined quality criteria. The success of a cooperation should not be measured solely by the number of views, but also by quality signals, such as saved posts or clicks on further information. Example: Former German Chancellor, Olaf Scholz, and successful TikTok influencer, Brooklyn, recorded up to around nine million views with their joint videos during the 2025 federal election campaign. 12. Embrace and own trends TikTok in particular, but also Instagram Reels, thrive on trends and challenges. However, trends are not an end in themselves. Selective use makes sense if the trend fits the specific profile and can be filled with content. But don‘t just copy pop culture trends – experimental formats can also come from politics itself. It can help to screen trends regularly, personalise only a few of them appropriately and always add a peculiar twist. If a trend is clearly unsuitable, it should be consistently discontinued. Short tests with small target groups help to avoid reputational risks. Example: Left-wing Member of the Bundestag, Caren Lay, raps about the need for a rent cap in her ‚Rhubarb Song‘; and law student and activist, Theresia Crone, uses fashion as a gateway to educate young people about current political debates with her ‚political outfit check‘. 13. Substance over reach Presence on video platforms is essential, but algorithms reward simplification, polarisation and escalation. Political communication must not be driven by this. Formats serve as a bridge for conveying democratic values and concrete solutions; long-term credibility outweighs short-term effects. Before posting, it is important to go through the content checklist, which includes the core message, evidence, relevance and the next step. Door-opener clips should be followed by in-depth explanations of political issues, supported by reliable sources and further reading. It is also advisable to define a clear-cut blend of targets and posts that prioritises information content and only uses easy entry points in a targeted manner. The entertaining element remains central on short video platforms – even if it only serves as a hook. November 2025How to Sell Democracy Online (Fast) 5555 Recommendations for people who talk to young people about politics and media use It is not only political actors who can contribute to the sensitive communication of political information and a pluralistic exchange of opinions between democratic forces and young people. Young people themselves can also reflect on their media usage behaviour, learn to take a critical look at political content and be encouraged to participate more actively. People who talk to young people about politics and media usage (such as parents, educators or even political influencers) can ask about the needs and interests of the younger generation and also point out potentially problematic conditions or uses. For example: • Instagram and TikTok are not public, non-profit spaces, but private, profit-oriented platforms. • Social media platforms offer low-threshold opportunities for direct political exchange with elected representatives and fellow citizens. A pluralistic and constructive use of these platforms can be conducive to democratic discourse. • Content directly from political actors or influencers has no journalistic contextualisation, i.e. no critical classification. • Personal feeds are algorithmically controlled, meaning that certain content is displayed prominently while other content is not. Unless one actively searches for reputable news sources, there is a reliance on the ”news find me” effect, which is algorithmically selected, incomplete and biased. • Influencers do not necessarily have political or journalistic training and may also post content on behalf of certain political parties or organisations. • And much more. 6. ”How to sell” in a nutshell Dos Direct address Eye contact: Look into the camera and talk to the audience in an informal manner. In a few authentic cases, a more formal form of address also works. Target group clarity Instagram and TikTok have different target groups (e.g. age structure) – one size does not fit all. Address Clear messages • Good organisation: What is the core message to be conveyed? What information is needed to do this? • Comprehensibility: Short sentences, few technical terms or foreign words + use subtitles • Storyline: Make the content steps (problem, goal, solution) comprehensible, e.g. through cuts. • Evidence: For increased complexity, use source citation, statistics and references. Content Online action with offline impact Calls for offline action are well received. The same applies to calls to vote. Young people rarely participate in online discussions. Encouraging calls to action can increase interaction. Call to action & interaction 5756 Political communication, especially across the broad spectrum from press relations to social media, is time-consuming. Not all elected representatives have the resources and staff to do this. Below, we attempt to summarise the dos and don‘ts for short political videos that can be produced without requiring significant additional resources. Don‘ts Addressing young people ”Hey adolescents!” would not be a good way to start a video. Addressing a group is no substitute for an approachable, empathetic and understanding tone of voice. Address Nothing artificial • No extremes: No exclusively positive acclaims (about oneself or one‘s party), no personalised, disproportionate attacks on political opponents. • No script: Unnatural, curated scenes or dialogues are recognised and disapproved of. • No cooking instructions: Purely apolitical content from political actors is frowned upon. Content Short and quick Right at the beginning: Set the hook (3 seconds) and, if necessary, name the topic; make slow speakers speak faster (increase the tempo with abrupt cuts, if necessary), but keep it understandable. Selfies and affordances Videos can be recorded in selfie mode as long as the sound and image quality do not suffer. Filters and/or effects can be tested sparingly. Trends and humour Both may be used occasionally, but not as a substitute for substantive content. A political reference is necessary, even if the focus of the video is not on political issues. Approachability Young people want politicians to show their human side – e.g. through dialogue, interaction and an honest aura. Recognition value The short videos should develop their own style and thus recognisable features, e.g. through the exact same introduction, listicles, stiches or theme. Style Complicated and artificial language Long-winded, abstract and complex: neither liked nor understood. Forced attempts to sound young (using youth language or words) do not work either. Condescending Young people do not appreciate condescending behaviour towards others and superficial assertions, neither of which are conducive to democratic dialogue. Filters and effects While filters are not necessary, good sound and a stable image are: avoid poor recording quality. Style No like-begging Online calls to action, such as liking or sharing, are frowned upon. Call to Action & Interaction 57 November 2025How to Sell Democracy Online (Fast) 5858 Dr. Pablo Jost is a communication scientist at the Institute for Journalism Studies at Johannes Gutenberg University Mainz, where he completed his doctorate in 2022 on ”Popularity indicators in political communication research”. He currently researches and teaches as a visiting professor at the Institute for Journalism and Communication Research at the Hanover School of Music, Drama and Media. As co-founder and strategic advisor to the Federal Association for Countering Online Hate, he investigates the mobilization of radical and extremist protest movements on digital platforms and their offline effects. His research interests also include the media representation of social controversies, the communication of political actors and their adaptation to the conditions of digitalization. Paulina Fröhlich is deputy managing director and responsible for the focus ”Resilient Democracy” at the Berlin think tank Das Progressive Zentrum. There, she heads the democracy conference ”lnnocracy”, designs dialog and discourse spaces and is co-author of several publications such as ”Die Übergangenen - Strukturschwach und Erfahrungsstark.” (2022). After completing her studies, Paulina co-founded the Kleiner Fünf initiative, which supports people in taking action against right-wing populism with the help of ”radical politeness”. She comments on political events for Deutschlandfunk Kultur and regularly on ”Die Frage des Tages” at Radio3. In the winter semester 22/23, Paulina taught as a guest lecturer at the University of Arts in Berlin and was appointed to the Forum #Zukunftsstrategie at the German Federal Ministry of Education and Research in 2023. The Authors and the Team Melanie Weiser is a project manager for democratic policy projects in the focus area of ”Resilient Democracy” at the Berlin think tank Das Progressive Zentrum. Previously, she worked in the German Bundestag at the interface of the Ministry of Economic Affairs and Climate Protection and at the Alliance for Securing Democracy of the German Marshall Fund in Brussels. She completed her studies in the fields of politics and international relations as well as international security at Lake Constance, in Berlin and in the USA. November 2025How to Sell Democracy Online (Fast) 6565 Cartes-Barroso, M. J., García-Estévez, N., & Méndez-Muros, S. (2025). Attracting the Vote on TikTok: Far-Right Parties’ Emotional Communication Strategies in the 2024 European Elections. Journalism and Media, 6(1), 33. https://doi.org/10.3390/ journalmedia6010033 Cazzamatta, R. (2025, Februar 24). Das Ende von Metas Fact-Checking-Programm: Warum Community Notes professionelle Faktenchecks nicht ersetzen können. Universität Erfurt Forschungsblog Wortmelder. https://www.uni-erfurt.de/forschung/aktuelles/forschungsblog-wortmelder/ das-ende-von-metas-fact-checking-programm Chang, H.-C. H., Shaman, B., Chen, Y.-C., Zha, M., Noh, S., Wei, C., Weener, T., & Magee, M. (2024). Generative Memesis: AI Mediates Political Information in the 2024 United States Presidential Election. https://doi. org/10.2139/ssrn.5007096 Chu, Y., Denemark, D., Denemark, D., Donovan, T., Fuh-sheng Hsieh, J., Gao, E., Gunther, R., Huang, M., Lagos, M., Lin, J., Mattes, R., Mattes, R., Matukhno, N., Mishler, W., Moreno, A., Niemi, R. G., Niemi, R. G., Palshikar, S., Ramón Montero, J., ... Yongrong, C. (Hrsg.). (2022). Growing Up Democratic: Does It Make a Difference? Lynne Rienner Publishers. https://doi. org/10.1515/9781626375567 Classen, K., Kollmer, A., Schlage, M., Schöpflin, A., Winkler, J., & Witerspan, H. (2024). Right-wing populist communication of the party AfD on TikTok: To what extend does the AfD use TikTok as part of its communication to win over young voters? In A. Godulla, C. Buller, V. Freudl, I. Merz, J. Twittenhoff, J. Winkler, & L. Zapke (Hrsg.), The Dynamics of Digital Influence: Communication Trends in Business, Politics and Activism (S. 100–122). Collazos Garcia, R. (2024, Januar 26). Das sind die Social-Media-Trends 2024. absatzwirtschaft. https://www.absatzwirtschaft.de/ das-sind-die-social-media-trends-2024-254260/ Cremers, K., Goyn, L., Krüper, T., Lehmann, C., & Liehr, A. (2022). Cheapest Is dearest, though far from pro - fessional: A qualitative study on the use of social media during the Federal Election 2021 in Germany. In A. Godulla, L. Beck, E. Christiansen, P. A. Johe, T. Krüper, V. Niemsch, & F. Saxinger (Hrsg.), Disrupt Adapt: New ways to deal with current challenges in media and communication (S. 43–71). Decker, F. (2022, Dezember 2). Etappen der Parteigeschichte der AfD. Bundeszentrale für politische Bildung. Parteien in Deutschland. https://www.bpb. de/themen/parteien/parteien-in-deutschland/ afd/273130/etappender-parteigeschichte-der-afd/ Decker, F. (2025). Bundestagswahl 2025 (Informationen zur politischen Bildung / izpb). https:// www.bpb.de/shop/zeitschriften/izpb/558394/ bundestagswahl-2025/ Demografieportal. (2025). Altersspezifische Wahlbeteiligung. https://www.demografie-portal.de/DE/Fakten/ wahlbeteiligung.html Deutsche Kinderund Jugendstiftung. (2025, Februar 24). Politische Einstellungen und Wahlverhalten junger Menschen. Deutsche Kinderund Jugendstiftung GmbH. https://www.dkjs.de/wp-content/ uploads/2025/02/analyse-politische-einstellungenund-wahlverhalten-junger-menschen.pdf Deutscher Bundesjugendring. (2021, November 12). Jugendverbände als politische Akteure. Deutscher Bundesjugendring. https://www.dbjr.de/artikel/ jugendverbaende-als-politische-akteure Deutscher Städtetag. (2025). Die Bedeutung von Social Media für die Kommunikationsarbeit der Städte. Positionspapier des Deutschen Städtetages. staedtetag.de. https://www.staedtetag.de/files/dst/docs/Publikationen/Positionspapiere/2025/positionspapier-socialmedia-kommunikationsarbeit.pdf November 2025How to Sell Democracy Online (Fast) 6666 DeVito, M. A. (2017). From Editors to Algorithms: A values-based approach to understanding story selection in the Facebook news feed. Digital Journalism, 5(6), 753–773. https://doi.org/10.1080/216708 11.2016.1178592 Dienlin, T., & Johannes, N. (2020). The impact of digital technology use on adolescent well-being. Dialogues in Clinical Neuroscience, 22(2), 135–142. https://doi. org/10.31887/DCNS.2020.22.2/tdienlin Dodds, D. (2024, März 19). Short-Form Video Content: Capturing Attention In The Digital Age. Forbes. https://www.forbes.com/councils/ forbesagencycouncil/2024/03/19/short-form-video -content-capturing-attention-in-the-digital-age/ Doroshenko, L., & Tu, F. (2023). Like, Share, Comment, and Repeat: Far-right Messages, Emotions, and Amplification in Social Media. Journal of Information Technology & Politics, 20(3), 286–302. https://doi.org /10.1080/19331681.2022.2097358 Edwards, A. (1957). The social desirability variable in personality assessment and research. Dryden Press. Engesser, S., Ernst, N., Esser, F., & Büchel, F. (2017). Populism and social media: How politicians spread a fragmented ideology. Information, Communication & Society, 20(8), 1109–1126. https://doi.org/10.1080/ 1369118X.2016.1207697 European Commission. (2024, April 30). Press Release. Commission opens formal proceedings against Facebook and Instagram under the Digital Services Act. digital-strategy.ec.europa.eu. https://digital-strategy. ec.europa.eu/en/news/commission-opens-formalproceedings-against-facebook-and-instagram-underdigital-services-act European Parliament & Ipsos European Public Affairs (Hrsg.). (2023). Media & News survey 2023. European Parliament. https://doi.org/10.2861/11595 European Union. (2025). Fact-checking and Content Moderation. AT A GLANCE. Digital issues in focus. https://www.europarl.europa.eu/RegData/etudes/ ATAG/2025/769493/EPRS_ATA(2025)769493_EN.pdf European Youth Portal. (2024). Empowering Europe’s Youth: Volt Europa’s Innovative Approach to Political Engagement. European Youth Portal. https://youth. europa.eu/news/empowering-europes-youth-volteuropas-innovative-approach-political-engagement_ en Feierabend, S., Rathgeb, T., Gerigk, & Glöckler, S. (2024). JIM 2024. Jugend, Information Medien. Basis - untersuchung zum Medienumgang 12bis 19-jähriger in Deutschland. Medienpädagogischer Forschungs - verbund Südwest. Fielitz, M., Sick, H., Schmidt, M., & Donner, C. (2024). Social-Media-Partei AfD? Digitales Landtagswahlkämpfe im Vergleich (OBS-Arbeitspapier 73). Otto Brenner Stiftung. https://www.otto-brenner-stiftung.de/ social-media-partei-afd/ Friedrich-Naumann-Stiftung. (2025, März 18). Desinformation. Umfrage: Desinformation unter jungen Menschen weit verbreitet. freiheit.org. https://www.freiheit. org/de/deutschland/umfrage-desinformation-unterjungen-menschen-weit-verbreitet Geise, S., Maubach, K., & Boettcher Eli, A. (2024). Picture me in person: Personalization and emotionalization as political campaign strategies on social media in the German federal election period 2021. New Media & Society, 14614448231224031. https:// doi.org/10.1177/14614448231224031 Gil De Zúñiga, H., Weeks, B., & Ardèvol-Abreu, A. (2017). Effects of the News-Finds-Me Perception in Communication: Social Media Use Implications for News Seeking and Learning About Politics. Journal of Computer-Mediated Communication, 22(3), 105–123. https://doi.org/10.1111/jcc4.12185 November 2025How to Sell Democracy Online (Fast) 6767 Gille, M. (2025). Jugend und Politik. Die Auswirkungen des Wandels in der Sozialstruktur fortgeschrittener Industriegesellschaften auf die politischen Orientierungen Jugendlicher. Bundeszentrale für politische Bildung. Kurz & Knapp. https://www.bpb.de/kurz-knapp/lexika/ handwoerterbuch-politisches-system/202044/ jugend-und-politik/ Gonzalez, A., Schmuck, D., & Vandenbosch, L. (2024). Posting and framing politics: A content analysis of celebrities’, athletes’, and influencers’ Instagram political content. Information, Communication & Society, 27(8), 1605–1627. https://doi.org/10.1080/1369 118X.2023.2285495 Goodwin, A., Joseff, K., Riedl, M. J., Lukito, J., & Woolley, S. (2023). Political Relational Influencers: The Mobilization of Social Media Influencers in the Political Arena. International Journal of Communication, 17, 1613–1633. Görtz, R. von, & Langness, A. (2024). Jung. Kritisch. Demokratisch. – Perspektiven junger Erwachsener auf die Herausforderungen unserer Zeit. https://doi. org/10.11586/2024007 Guess, A. M., Malhotra, N., Pan, J., Barberá, P., Allcott, H., Brown, T., Crespo-Tenorio, A., Dimmery, D., Freelon, D., Gentzkow, M., González-Bailón, S., Kennedy, E., Kim, Y. M., Lazer, D., Moehler, D., Nyhan, B., Rivera, C. V., Settle, J., Thomas, D. R., ... Tucker, J. A. (2023). How do social media feed algorithms affect attitudes and behavior in an election campaign? Science, 381(6656), 398–404. https://doi.org/10.1126/science.abp9364 Harff, D., & Schmuck, D. (2025). Who Relies on Social Media Influencers for Political Information? A Cross-Country Study Among Youth. The International Journal of Press/Politics, 30(3), 841–864. https://doi. org/10.1177/19401612241234898 Haselmayer, M. (2019). Negative campaigning and its consequences: A review and a look ahead. French Politics, 17(3), 355–372. https://doi.org/10.1057/ s41253-019-00084-8 Haßler, J. (2024, März 1). Soziale Medien und die Kommunikation politischer und staatlicher Institutionen. Bundeszentrale für Politische Bildung. Soziale Medien. https://www.bpb.de/themen/medien-journalismus/ soziale-medien/545485/sozialemedien-und-diekommunikation-politischer-und-staatlicher-institutionen/ Haßler, J., Kümpel, A. S., & Keller, J. (2023). Instagram and political campaigning in the 2017 German federal election. A quantitative content analysis of German top politicians’ and parliamentary parties’ posts. Information, Communication & Society, 26(3), 530–550. https://doi.org/10.1080/1369118X.2021.1954974 Haßler, J., Wurst, A.-K., & Pohl, K. (2024). Politicians over issues? Visual personalization in three Instagram election campaigns. Information, Communication & Society, 27(5), 815–835. https://doi.org/10.1080/136 9118X.2023.2227684 Hillje, J. (2024). Social Media: Die digitale Dominanz der AfD brechen! Blätter für deutsche und internationale Politik, 13–16. Hillje, J. (2025). Mehr Emotionen wagen: Wie wir Angst, Hoffnung und Wut nicht dem Populismus überlassen. Piper. Höhne, V. (2024, Juni 27). Hat der Hashtag #ReclaimTiktok funktioniert? Süddeutsche Zeitung Dossier. https:// www.sz-dossier.de/dossiers/platz-der-republik/202406-27-hat-der-hashtag-reclaimtiktok-funktioniert Hohner, J., Kakavand, A. E., & Rothut, S. (2024). Analyzing radical visuals at scale: How far-right groups mobilize on TikTok. Journal of Digital Social Research, 6(1). https://doi.org/10.33621/jdsr.v6i1.200 Hügelmann, B. (2021). Political Influencers? How voters make up their minds in the age of social media and digital communications. https://doi.org/10.13140/ RG.2.2.30712.93443 November 2025How to Sell Democracy Online (Fast) 6868 Hulbert, D. (2023). dave1010/tree-of-thought-prompting: First version (Version v0.1) [Software]. Zenodo. https://doi.org/10.5281/ZENODO.10323452 Ibrahim, H., Jang, H. D., Aldahoul, N., Kaufman, A. R., Rahwan, T., & Zaki, Y. (2025). TikTok’s recommendations skewed towards Republican content during the 2024 U.S. presidential race (Version 2). arXiv. https://doi. org/10.48550/ARXIV.2501.17831 Instagram. (2024, Februar 9). Ankündigungen. Wir weiten unser Konzept zu politischen Inhalten auf Instagram und Threads aus. About Instagram Blog. https:// about.instagram.com/de-de/blog/announcements/ continuing-our-approach-to-political-content-oninstagram-and-threads Instagram. (2025, August 6). Ankündigungen. Neue Funktionen, mit denen du dich mit Freund*innen verbinden kannst. About Instagram Blog. https:// about.instagram.com/blog/announcements/ instagram-new-features-connect-friends/ Institut für Generationenforschung. (2025). Jugendwahlstudie 2025. Analyse der Erstwählerkohorte in Bezug auf die Bundestagswahl 2025. Institut für Generationenforschung. https://www.generation-thinking.de/ post/jugendwahlstudie-2025 Jolly, S., Bakker, R., Hooghe, L., Marks, G., Polk, J., Rovny, J., Steenbergen, M., & Vachudova, M. A. (2022). Chapel Hill Expert Survey trend file, 1999–2019. Electoral Studies, 75, 102420. https://doi.org/10.1016/j. electstud.2021.102420 Jost, P. (2023). How politicians adapt to new media logic. A longitudinal perspective on accommodation to user-engagement on Facebook. Journal of Information Technology & Politics, 20(2), 184–197. https://doi. org/10.1080/19331681.2022.2076271 Jost, P., Fröhlich, P., & Weiser, M. (2025). POV: Wahlkampf. Eine Analyse politischer Kurzvideos auf TikTok und Instagram im Bundestagswahlkampf 2025. Bertelsmann Stiftung. Das Progressive Zentrum. Jost, P., Heft, A., Buehling, K., Zehring, M., Schulze, H., Bitzmann, H., & Domahidi, E. (2023). Mapping a Dark Space: Challenges in Sampling and Classifying Non-Institutionalized Actors on Telegram. Medien & Kommunikationswissenschaft, 71(3–4), 212–229. https://doi.org/10.5771/1615-634X-2023-3-4-212 Jost, P., Maurer, M., & Hassler, J. (2020). Populism Fuels Love and Anger: The Impact of Message Features on Users’ Reactions on Facebook. International Journal of Communication, 14, 2081–2102. Kaack, H. (1971). Geschichte und Struktur des Deutschen Parteiensystems. VS Verlag fur Sozialwissenschaften GmbH. Katzy-Reinshagen, A., Degeling, M., Barth, S., & Dorn, M. (2025). Wahlkampf im Feed? Wie TikTok mit parteipolitischen Inhalten im Vorfeld der Bundestagswahl 2025 umgeht. https://isdgermany.org/wp-content/ uploads/2025/02/ISD-Analyse-Wahlkampf-im-Feed. pdf Khamis, S., Ang, L., & Welling, R. (2017). Self-branding, ‘micro-celebrity’ and the rise of Social Media Influencers. Celebrity Studies, 8(2), 191–208. https://doi.org/ 10.1080/19392397.2016.1218292 Kligler-Vilenchik, N., & Literat, I. (2024). Not your parents’ politics: Understanding young people’s political expression on social media. Oxford University Press. Knodel, K. (2023). A party-typological classification of Volt Europa. Between digital party and movement party? https://doi.org/10.13140/RG.2.2.12140.44163 November 2025How to Sell Democracy Online (Fast) 6969 Koenigsdorff, S., Plavec, J. G., & Wolf, J. (2025, Mai 15). Politik auf TikTok. Bislang größte Nutzerstudie zu TikTok – Die Ergebnisse. https://www.stuttgarter-zeitung.de/ inhalt.politik-auf-tiktok-bislang-groesste-nutzerstudie-zu-tiktok-die-ergebnisse.7922105b-2f6f-4c0dbc8f-7ded3776e1f7.html Kümpel, A. S., & Rieger, D. (2020). Kann Instagram auch Politik? Beeinflussung der Informationsnutzung und Effekte für die Meinungsbildung durch politische Inhalte. Konrad-Adenauer-Stiftung e.V. Lee, S., Park, S., & Lee, J. (2025). Exploiting FineGrained Skip Behaviors for Micro-Video Recommendation (Version 1). arXiv. https://doi.org/10.48550/ ARXIV.2504.03107 Lou, C., & Kim, H. K. (2019). Fancying the New Rich and Famous? Explicating the Roles of Influencer Content, Credibility, and Parental Mediation in Adolescents’ Parasocial Relationship, Materialism, and Purchase Intentions. Frontiers in Psychology, 10, 2567. https:// doi.org/10.3389/fpsyg.2019.02567 Lübcke, S. (2022). Authenticity in Political Communication: Conceptualization, Measurement, and Examination of a Popular Concept [Dissertation, Friedrich-SchillerUniversität Jena]. https://www.db-thueringen.de/ receive/dbt_mods_00053790 Magin, M., Podschuweit, N., Haßler, J., & Russmann, U. (2017). Campaigning in the fourth age of political communication. A multi-method study on the use of Facebook by German and Austrian parties in the 2013 national election campaigns. Information, Communication & Society, 20(11), 1698–1719. https://doi. org/10.1080/1369118X.2016.1254269 Manemann, T., & Lehmann, T. (2021, Juli 13). Wahlkampf als Unterhaltung. Die AfD läuft sich auf TikTok warm. Belltower News. https://www.belltower.news/ wahlkampf-als-unterhaltung-die-afd-laeuft-sich-auftiktok-warm-118465/ McDonnell, S., Döbele, C., Feldmann, D., & von Görtz, R. (2024). Junges Engagement für sozialen Wandel: Ergebnisse einer Befragung junger Menschen zu politischen Einstellungen und politischem Engagement. https://doi. org/10.11586/2024171 Medina Serrano, J. C., Papakyriakopoulos, O., & Hegelich, S. (2020). Dancing to the Partisan Beat: A First Analysis of Political Communication on TikTok. 12th ACM Conference on Web Science, 257–266. https://doi. org/10.1145/3394231.3397916 Meijers, M. J., & Zaslove, A. (2021). Measuring Populism in Political Parties: Appraisal of a New Approach. Comparative Political Studies, 54(2), 372–407. https:// doi.org/10.1177/0010414020938081 Meta. (2024, November 28). Meta’s Progress Implementing the Digital Services Act. Meta. https://about.fb.com/ news/2024/11/metas-progress-implementing-thedigital-services-act/ Meta Platforms, Inc. (2024, April 24). First Quarter 2024 Results Conference Call. Protocol. https://s21. q4cdn.com/399680738/files/doc_financials/2024/ q1/META-Q1-2024-Earnings-Call-Transcript.pdf Metzler, H., & Garcia, D. (2024). Social Drivers and Algorithmic Mechanisms on Digital Media. Perspectives on Psychological Science, 19(5), 735–748. https://doi. org/10.1177/17456916231185057 Meyer, H., Niemann-Lenz, J., Rodeck, L., & Revers, M. (2024). Beyond Anti-Elitism and Out-Group Attacks: How Concerns Shape the AfD’s Populist Representation on German TikTok During the 2024 European Elections. Center for Open Science. https://doi.org/10.31219/ osf.io/yk3u4 Moir, A. (2023). The Use of TikTok for Political Campaigning in Canada: The Case of Jagmeet Singh. Social Media + Society, 9(1). https://doi. org/10.1177/20563051231157604 November 2025How to Sell Democracy Online (Fast) 7070 Molem, A., Makri, S., & Mckay, D. (2024). Keepin’ it Reel: Investigating how Short Videos on TikTok and Instagram Reels Influence View Change. Proceedings of the 2024 ACM SIGIR Conference on Human Information Interaction and Retrieval, 317–327. https://doi. org/10.1145/3627508.3638341 Mudde, C. (2004). The Populist Zeitgeist. Government and Opposition, 39(4), 541–563. https://doi. org/10.1111/j.1477-7053.2004.00135.x Müller, T. (2024). Ergebnisse der ARD/ZDF-Medienstudie 2024. Zahl der Social-Media-Nutzenden steigt auf 60 Prozent. Media Perspektiven, 28. Netzschreier. (2025). Sechs Monate nach der Wahl: Wer im Bundestag die sozialen Medien dominiert. Netzschreier Presse. https://netzschreier.com/presse/ neuer-bundestag-social-media-vergleich/ Newman, N., Ross Arguedas, A., Robertson, C. T., Nielsen, R. K., & Fletcher, R. (2025). Reuters Insitute Digital News Report 2025. Reuters Institute for the Study of Journalism. https://doi.org/10.60625/RISJ-8QQF-JT36 NYU Tandon. (2022, Oktober 21). Press Release. Facebook and TikTok failing to prevent spreading of blatant US midterms disinformation. NYU Tandon School of Engineering. https://engineering.nyu.edu/news/facebook-and-tiktok-failing-prevent-spreading-blatantus-midterms-disinformation Okroi, J., & Wendt, F. (2024). VoiceUp! Junge Stimmen zu Meinungsfreiheit. Auswertungsbericht von qualitativen Gruppendiskussionen der Initiative VoiceUp! Im Jahr 2024. Deutsche Kinderund Jugendstiftung GmbH. Orben, A., & Przybylski, A. K. (2019). Screens, Teens, and Psychological Well-Being: Evidence From Three Time-Use-Diary Studies. Psychological Science, 30(5), 682–696. https://doi.org/10.1177/0956797619830329 Petrocik, J. R. (1996). Issue Ownership in Presidential Elections, with a 1980 Case Study. American Journal of Political Science, 40(3), 825–850. https://doi. org/10.2307/2111797 Polborn, M. K. (2006). Informative Positive and Negative Campaigning. Quarterly Journal of Political Science, 1(4), 351–371. https://doi.org/10.1561/100.00000013 Presseund Informationsamt der Bundesregierung. (2025, Januar 22). Politisches Informationsverhalten. Ergebnisse aus der Meinungsforschung. Table Briefings. https://table.media/wp-content/ uploads/2025/01/23162309/Informationsverhalten_BPA.pdf Quick, M., & Maddox, J. (2024). Us, Them, Right, Wrong: How TikTok’s Green Screen, Duet, and Stitch help shape political discourse. First Monday. https:// doi.org/10.5210/fm.v29i3.13360 Richter, C., Ebert, M., Rau, J., & Kessling, P. (2025). Wahlkampf im Takt des Algorithmus: Warum die AfD auf TikTok erfolgreich ist und was demokratische Parteien besser machen können. Forschungsinstitut Gesellschaftlicher Zusammenhalt. https://fgzrisc.hypotheses.org/5898 Riedl, J., Steup, J., Kielbassa, P., Drews, W., Müller, A., Neumeier, A., Dafnos, A., Shukla, C., Wiliams, T., & Fischer-Kattner, A. (2025). Social Media Monitoring zur Bundestagswahl 2025. SPARTA. https://dtecbw. de/sparta/ Rooduijn, M., Pirro, A. L. P., Halikiopoulou, D., Froio, C., van Kessel, S., de Lange, S. L., Mudde, C., & Taggart, P. (2023). The PopuList 3.0: An Overview of Populist, Farleft and Far-right Parties in Europe. [Dataset]. https:// doi.org/10.17605/OSF.IO/2EWKQ Rovira Kaltwasser, C. (2024). The transformation of the mainstream right in Western Europe [Policy Brief]. Foundation for European Progressive Studies. https:// feps-europe.eu/publication/the-transformation-ofthe-mainstream-right-in-western-europe/ November 2025How to Sell Democracy Online (Fast) 7171 Salganik, M. J. (2019). Bit by bit: Social research in the digital age. Princeton University Press. Scherndl, G., & Nicolaus, K. (2024, August 22). Organi - sierter Stimmenfang auf Tiktok: Verurteilter Volksverhetzer wirbt für die AfD. CORRECTIV. https://correctiv.org/ faktencheck/hintergrund/2024/08/22/rechte-propaganda-afd-werbung-von-tiktok-guerilla-shlomofinkelstein-angerverse/ Schnabel, D., & Berendsen, E. (2024). Das TikTokUniversum der (extremen) Rechten – Trends, Strategien und Ästhetik in der Social Media-Kommunikation. Bildungsstätte Anne Frank. https://www. bs-anne-frank.de/mediathek/publikationen/ das-tiktok-universum-der-extremen-rechten Schnetzer, S., Hampel, K., & Hurrelmann, K. (2024). Trendstudie Jugend in Deutschland 2024: Verantwortung für die Zukunft? Ja, aber. https://simon-schnetzer.com/ trendstudie-jugend-in-deutschland-2024/ Schubert, K. (2024, Juli 24). Union gewinnt – AfD Nummer zwei. zdf heute. https://www.zdfheute.de/ politik/europawahl-2024-ergebnisse-deutschlandeuropa-100.html Sekwenz, M.-T., Wagner, B., & Bruijn, H. D. (2025). From Reports to Reality: Testing Consistency in Instagram’s Digital Services Act Compliance Data. arXiv. https://doi. org/10.48550/arXiv.2507.01787 Sick, H., Fielitz, M., Donner, C., Schmidt, M., Marcks, H., & Brodersen, W. (2024). Subscribe to Subversion! Crossmediale Techniken der Gemeinschaftsbildung in rechtsalternativen Kontexten. Machine Against the Rage, 2(2). https://doi.org/10.58668/matr/06.2 Song, H., Gil De Zúñiga, H., & Boomgaarden, H. G. (2020). Social Media News Use and Political Cynicism: Differential Pathways Through ”News Finds Me” Perception. Mass Communication and Society, 23(1), 47–70. https://doi.org/10.1080/15205436.2019.1651867 Stattin, H., Amnå, E., & Russo, S. (2023). Setting societal engagement goals during adolescence amplifies the impacts of political interest on political activities during young adulthood. International Journal of Behavioral Development, 47(2), 135–145. https://doi. org/10.1177/01650254221113465 Steinhauer, R., Stolz, H.-J., Moldenhauer, S., & Rüther, E. (2025). Bedarfe von Kindern und Jugendlichen für ein gelingendes Aufwachsen: Eine bundesweite Erhebung von monetären und bildungsbezogenen Bedarfen junger Menschen. https://doi.org/10.11586/2025031 Stier, S., Bleier, A., Bonart, M., Mörsheim, F., Bohlouli, M., Nizhegorodov, M., Posch, L., Maier, J., Rothmund, T., & Staab, S. (2018). Systematically Monitoring Social Media: The case of the German federal election 2017 (Bd. 2018/04). GESIS – Leibniz-Institut für Sozialwissenschaften. https://doi.org/10.21241/ssoar.56149 Stier, S., Bleier, A., Lietz, H., & Strohmaier, M. (2018). Election Campaigning on Social Media: Politicians, Audiences, and the Mediation of Political Communication on Facebook and Twitter. Political Communication, 35(1), 50–74. https://doi.org/10.1080/1058 4609.2017.1334728 Sumpter, D. J. T. (2021). Outnumbered: From Facebook and Google to fake news and filter-bubbles -- the algorithms that control our lives (First edition). Zed Books. https://doi.org/10.5040/9781472947451 Suuronen, A., Reinikainen, H., Borchers, N. S., & Strandberg, K. (2022). When Social Media Influencers Go Political: An Exploratory Analysis on the Emergence of Political Topics Among Finnish Influencers. Javnost – The Public, 29(3), 301–317. https://doi.org/1 0.1080/13183222.2021.1983367 tagesschau.de. (2024, April 10). Die Presenterinnen und Presenter der @tagesschau. tagesschau. https://www. tagesschau.de/ueber-uns/moderatoren-sprecher/ tagesschau-social-media-100.html November 2025How to Sell Democracy Online (Fast) 7272 tagesschau.de. (2025a). Bundestagswahl 2025. Stimmanteile bei Erstwählenden im Vergleich zu 2021. tagesschau. https://www.tagesschau.de/wahl/ archiv/2025-02-23-BT-DE/charts/umfrage-werwas/ chart_1874859.shtml tagesschau.de. (2025b, Juli 26). Bündnis Sahra Wagenknecht. Jugendverband des BSW gegründet. tagesschau. https://www.tagesschau.de/inland/innenpolitik/bswjugendorganisation-100.html Thomeczek, J. P. (2024). Bündnis Sahra Wagenknecht (BSW): Left-Wing Authoritarian – And Populist? An Empirical Analysis. Politische Vierteljahresschrift, 65, 535–552. Thorson, K., & Wells, C. (2016). Curated Flows: A Framework for Mapping Media Exposure in the Digital Age: Curated Flows. Communication Theory, 26(3), 309–328. https://doi.org/10.1111/comt.12087 TikTok. (2023). DSA RISK ASSESSMENT REPORT 2023. https://sf16-va.tiktokcdn.com/obj/eden-va2/zayvwlY_ fjulyhwzuhy%5b/ljhwZthlaukjlkulzlp/DSA_H2_2024/ TikTok-DSA-Risk-Assessment-Report-2023.pdf TikTok for Business. (2022, November 20). TikTok Works: How entertainment on TikTok improves efficiency for brands. TikTok for Business Blog. https://ads. tiktok.com/business/en-SG/blog/tiktok-works-howentertainment-improves-efficiency-for-brands Tjaden, J., Wolfgram, J., Philipp, A., Weissmann, S., Bobzien, L., Kohler, U., & Verwiebe, R. (2024). Automated election audits – Analyzing Exposure to Political Content on Social Media with a Case Study of TikTok in Germany’s 2024 Regional Elections. https://doi.org/10.31219/osf. io/qwrjh Trujillo, A., Fagni, T., & Cresci, S. (2025). The DSA Transparency Database: Auditing Self-reported Moderation Actions by Social Media. Proceedings of the ACM on Human-Computer Interaction, 9(2), 1–28. https://doi. org/10.1145/3711085 TUI Stiftung. (2025). Junges Europa 2025. So denken Menschen zwischen 16 und 26 Jahren (Die Jugendstudie der TUI Stiftung). https://www.tui-stiftung.de/wp-content/uploads/2025/07/2025_0630_YouGov_Ergebnisbericht_TUI_Stiftung_Junges_Europa_2025_DE_final. pdf Vetter, A. (1997). Political Efficacy – Reliabilität und Validität: Alte und neue Meßmodelle im Vergleich. Deutscher Universitätsverlag. https://doi. org/10.1007/978-3-322-95354-4 Vodafone Stiftung. (2024). Junge Stimmen, starke Wirkung? Wie Jugendliche Freiheit in unserer Gesellschaft erleben. https://www.vodafone-stiftung.de/ wp-content/uploads/2024/05/VodafoneStiftungJunge-Stimmen-starke-Wirkung.pdf Weber, S. (2024, Dezember 5). Exklusive SWR-Umfrage. Bildung, Klima, Kriege und Zuwanderung – Das bewegt die Jugend in RLP. https://www.swr.de/swraktuell/ rheinland-pfalz/swr-umfrage-wie-tickt-die-jugendbildung-klima-krieg-100.html Weimann, G., & Masri, N. (2023). Research Note: Spreading Hate on TikTok. Studies in Conflict & Terrorism, 46(5), 752–765. https://doi.org/10.1080/10576 10x.2020.1780027 Wetzler, G., & Kiess, J. (2025, Februar 18). Wahlkampfarena TikTok. Die AfD verliert ihre Dominanz – zumindest offiziell. Belltower News. https://www.belltower. news/wahlkampfarena-tiktok-die-afd-verliert-ihredominanz-zumindest-offiziell-158355/ Wray-Lake, L., & Syvertsen, A. K. (2011). The developmental roots of social responsibility in childhood and adolescence. New Directions for Child and Adolescent Development, 2011(134), 11–25. https://doi. org/10.1002/cd.308 Wurst, A.-K., Pohl, K., & Haßler, J. (2023). Mobilization in the Context of Campaign Functions and Citizen Participation. Media and Communication, 11(3). https:// doi.org/10.17645/mac.v11i3.6660 November 2025How to Sell Democracy Online (Fast) 7373 Wurthmann, C., & Sältzer, M. (2025). Der Digitale Bundestagswahlkampf 2025: Erste Einblicke. Konrad Adenauer Stiftung. https://www.kas.de/de/web/ politische-bildung/politsnack/detail/-/content/derdigitale-bundestagswahlkampf-2025-erste-einblicke Zeng, J., Abidin, C., & Schäfer, M. S. (2021). Research Perspectives on TikTok & Its Legacy Apps| Research Perspectives on TikTok and Its Legacy Apps – Introduction. International Journal of Communication, 15(0), Article 0. S. 26 Fig. 1: Topics by Platform S. 27 Fig. 2: Topics of Actors by Party Affiliation S. 29 Fig. 3: Youth Reference by Actor and Party Affiliation S. 30 Fig. 4: Attacks and Acclaims by Actor S. 31 Fig. 5: Attacks and Acclaims by Party Affiliation S. 32 Fig. 6: Calls to Action by Actor S. 33 Fig. 7: Calls to Action by Party Affiliation S. 34 Fig. 8: Affordances by Actor S. 35 Fig. 9: Affordances by Party Affiliation S. 36 Fig. 10: Impact of Features on Number of Views S. 38 Fig. 11: Political Efficacy S. 39 Fig. 12: Places of Contact with Politics S. 40 Fig. 13: Use of Social Media S. 41 Fig. 14: Use of Political Contact by Politicians S. 42 Fig. 15: Use of Political Content by Influencers S. 44 Fig. 16: Opinion on Politicians’ Content on Social Media S. 45 Fig. 17: Desired Content from Politicians on Social Media S. 47 Fig. 18: Correlation Between Content and Selection S. 48 Fig. 19: Willingness to Engage in Subsequent Communication Zwick, D. (2025, Juli 9). Social-Media-Muffel. Einbruch bei Reichweite – der neuen Regierung fehlt es an StarInfluencern. WELT. https://www.welt.de/wirtschaft/ article256368176/Flaute-bei-X-Instagram-und-TikTokDer-neuen-Regierung-fehlt-es-an-Star-Influencern. html List of Figures November 2025How to Sell Democracy Online (Fast) 7474 Table A1: List of Influencer Accounts Account (Instagram/TikTok) #Videos TikTok #Videos Instagram .fuchsberger 151 0 Weichreitetv_ 1 0 afelia 0 4 aljosha_ aljosha__ 12 27 auf1.tv 0 150 aurelmertz 0 31 balaclava_graphics.bz 5 7 behamsarah 0 3 bjoern_banane 54 0 carlareemtsma 0 2 chef_maiki1 19 0 cult8877 124 0 daniela.sepehri/danielasepehri 15 128 demokrateam 0 11 derblauekanal 67 0 derloeffelbein 92 0 dianazurloewen 179 138 dievogtlaendischent 14 0 duzentekkal 0 123 elhotzo 0 2 evaschulz 6 20 fabiangrischkat 22 33 fraeulein_tessa 0 4 frank_thelen 0 42 frauloewenherz.de 15 22 h_lenah 42 0 hanneskreschel 88 0 hellipirelly 85 0 herr_anwalt/herranwalt 38 25 Appendix