Wanted: social media campaign support! A principal-agent study on the management of influencer collaborations by political parties in three EU countries

Social media has changed political communication, providing platforms where traditional and non-traditional voices can influence public opinion. Among these voices are social media influencers (hereafter: influencers). They are digital content creators who gather substantial followerships with which they build strong ties. Unlike established media channels, influencers operate primarily through social media platforms, leveraging their personal brand, audience engagement and digital reach to sway public discourse (Riedl et al., 2023). What is more, they increasingly perform allyship for political and social causes (Hund, 2023; Oliva et al., 2024; Wellman, 2022). Their rise highlights a transformative shift in how political messaging is crafted and disseminated as they facilitate their followers' understanding of political content and raise their interest in politics (Schmuck et al., 2022).

The appearance of influencers in the political arena comes at a time when political parties are facing the challenge of changing media consumption patterns among huge segments of the electorate (Ohme, 2019), searching for new approaches to strengthen their social media activities (Bracciale et al., 2021; Sneijder et al., 2018) and battling with declining voter turnout (Kostelka and Blais, 2021). Collaborations with influencers promise a remedy, as influencers can capture the attention of substantial social media audiences (Stubb et al., 2019) and even contribute to their voting behaviors (Pöyry and Reinikainen, 2024). Moreover, followers typically perceive influencer content as authentic, thus offering a solution to yet another considerable challenge political communicators struggle with (Luebke, 2021). But despite these promises, political communicators still appear to be more hesitant to embrace strategic influencer communication than their corporate counterparts.

This seeming hesitancy sparked our interest. In this study, we therefore set out to explore the perspective of political communicators – that is, party staff with communication management responsibilities – on influencer collaborations. More specifically, we assume that there are specific challenges that underlie the hesitancy to collaborate with influencers as well as solutions for dealing with these challenges. To identify challenges and solutions, we apply the principal-agent theory (PAT) (Rees, 1985a, b) because this theory provides a framework for systematically identifying the risks that collaborations entail.

In doing so, our study contributes to previous research in various ways. First, it broadens research on the management of influencer campaigns by investigating the political context. It thus complements previous research, which has focused on the corporate management of influencer collaborations. In particular, it adds to our understanding of obstacles to influencer collaborations and how communication managers attempt to overcome these obstacles. Second, by applying PAT, we address the current under-theorization of client-influencer relationships. Finally, by interviewing political communicators from three European Union countries – Finland, Germany and Romania – this study offers cross-country comparison in influencer research and, to our knowledge, the first such study on the management of influencer campaigns.

Influencers' ability to act as endorsers for commercial brands and impact consumer behaviors has been well documented in the literature on strategic communication and marketing (Reinikainen et al., 2020; Stubb et al., 2019; Wolf and Archer, 2018). As has become clear during past years, influencers are willing to not only discuss consumption choices, but also raise political and social issues (Suuronen et al., 2022). For instance, influencers have addressed issues such as racial justice (Wellman, 2022), climate politics (Allgaier, 2020) and taxation (Oliva et al., 2024). At the same time, pioneer research on the influence of influencers in the political arena has shown that many of the dynamics known from brand endorsements are also at play here. For example, followers are likely to be receptive to political messages by influencers who they find similar to themselves (Naderer, 2023) or who they have a parasocial relationship with (Cheng et al., 2024). Through these dynamics, influencers appear to be able to contribute to their followers' political interest (Schmuck et al., 2022) and even political participation (Dekoninck and Schmuck, 2022), such as voting (Pöyry and Reinikainen, 2024).

Such results imply that collaborating with influencers holds considerable potential also for political parties in addressing and winning over voters. However, while there are some examples of political parties utilizing influencers in their strategic communication, such as during the Harris/Walz 2024 presidential campaign (Kelly, 2024), many Western democracies have not witnessed an upsurge in the use of influencers in political campaigning.

Communication managers in political parties are apparently more hesitant to engage with influencers than their corporate counterparts. To understand how political communicators perceive collaborations with influencers and thus explain the reasons for their hesitancy, we suggest adopting a principal-agent perspective (Rees, 1985a, b). PAT aims at theorizing business relationships in which an agent performs functions on behalf of a principal. Originating in the field of economics, PAT has been applied for instance in the research on advertising (Koslow et al., 2022), public relations (Nothhaft and Stensson, 2019) and public affairs (Hoffmann et al., 2011). However, while authors have suggested also applying PAT to client–influencer relationships (Enke and Borchers, 2019; Sundermann and Raabe, 2019), this has, to our knowledge, not been done.

PAT holds that a principal (in our case: the political party) may decide to delegate the performance of particular functions to an agent (the influencer) because such a delegation can appear more effective than performing the functions themselves. However, this delegation comes at the price of information asymmetries. Information asymmetries exist when the principal does not have full insight into the qualities and doings of the agent. PAT assumes that the principal needs to manage these asymmetries so that the collaboration leads to optimal outcomes for the principal. Otherwise, the principal risks that the agent exploits the asymmetries in their favor. At its core, PAT therefore touches upon issues of balancing trust and control in relationships.

PAT distinguishes three asymmetries (e.g. Bernhold and Wiesweg, 2021; Saam, 2007; Steinle et al., 2014):

1. hidden characteristics: It is unclear for the principals what skills and abilities agents possess.

2. hidden intentions: It is unclear what motivations agents possess for agreeing to cooperate.

3. hidden actions: It is unclear (a) which effort agents invest and (b) whether they take the actions that are most expedient for collaboration success.

PAT locates these asymmetries in two different phases of collaboration. Communication managers in parties face the first two asymmetries when deciding with which influencers to collaborate (i.e. ex ante or before the collaboration), and they face the latter one during ongoing collaborations (i.e. ex post).

PAT has not been previously applied to the management of influencer campaigns. However, when revisiting existing research on the corporate management of influencer campaigns from a PAT perspective, it is possible to shed some light on how corporate communicators cope with the three information asymmetries. For a start, there exists research on the process of influencer selection for campaigns (Bakker, 2018; Borchers and Enke, 2021). Specifically, Borchers and Enke (2021) suggested grouping selection criteria into four groups, namely (1) fit and profile (e.g. thematic expertise), (2) influencer qualifications (e.g. storytelling skills), (3) social media indices (e.g. interaction rates) and (4) disqualifiers (e.g. collaborations with direct competitors). These findings provide insights into the characteristics and intentions communication managers want to see in influencers. Respective research also pointed out that in corporate communication, communication managers often collaborate with influencer agencies that organize the influencer selection process (Childers et al., 2019).

Furthermore, previous research has addressed the issue of control in influencer campaigns. The general thrust of research is that communication managers should and do refrain from controlling influencer conduct because any such control attempt will threaten the authenticity of influencers and, consequently, the persuasion success of the campaign (e.g. Childers et al., 2019; Leung et al., 2022; McMullan et al., 2022). However, Borchers (2023) challenged this view and presented several control techniques that communication managers developed to control influencer conduct despite a general paradigm of letting go of control.

Besides these two research interests – selection of influencers and control of influencer conduct – corporate approaches to cope with the information asymmetries suggested by PAT remain underexplored. Moreover, when focusing on the political context, we have virtually no insight into how party communicators navigate the risks of collaborating with influencers.

Adopting a new communication instrument, with which they have gained little or no experience, constitutes a risk for political communicators which they have to evaluate carefully. We argue that PAT offers a helpful analytical perspective for understanding their considerations because it provides a framework for organizing and distinguishing the various challenges that political communicators encounter when working with influencers. Following the information asymmetries theorized by PAT, we ask:

Since this study draws on data from three European countries, we add a comparative dimension to our analysis and ask:

This study is based on expert interviews with communication managers of parties in three European countries, Finland, Germany and Romania. As all three countries are members of the European Union, political parties have representatives in the same legislative body, the European parliament, and participate in the same elections every five years (the latest European parliament elections were held in May 2024). As EU countries, all three also fall under the same regulatory framework. All three countries have a multiparty system, and governments are typically formed with several parties.

In addition to the similarities, the three countries also have several differences among them. First, the countries differ in population size: Germany is the largest EU country with a population of 83.4 million, Romania a middle-sized EU country with a population 19.1 million people, and Finland a small EU country with a population of 5.6 million.

The countries have different democratic traditions. While Finland (1917) and Germany (1949) look back at long such traditions, Romania transformed into a democracy after the fall of the Iron Curtain (1991). Today, the Economist Democracy Index lists the first two countries as full democracies, whereas Romania qualifies as flawed democracy (Economist Intelligence, 2024). This history is reflected in the trust voters place in political institutions. Finland ranks the highest in both trust in the national parliament (59%) and political parties (34%), with Germany in the middle (43%/27%) and Romania the lowest (28%/24%) (Standard Eurobarometer, 2024).

The countries also vary in diffusion of social media. Finland has one of the highest shares of social media users among adult population in the EU (79%), with Romania close on its heeds (75%). Meanwhile, Germany ranks as the second to last in the EU (48%) (Statista, 2024). A somewhat similar pattern emerges when it comes to using social media as a news source. While Finland (43%) and Romania (48%) exceed the EU average of 37% of the population using social media for news consumption, Germany (28%) falls below it (European Union, 2023). Regarding preferred influencer content, the picture is again similar, although the differences are less pronounced. Here, 47% of Romanian social media users indicate they prefer political influencer content, compared to 40% in Finland and 35% in Germany (EU average: 38%) (European Union, 2023).

Finally, the three countries have different histories of influencer involvement in politics. In Germany, such involvement gained considerable attention relatively early, when the public attack of gaming YouTuber Rezo on the then-governing conservative party (CDU) for their climate politics became a major issue in media and politics in the run-up to the 2019 elections to the European Parliament (Allgaier, 2020). In Finland, influencers have been acknowledged as essential stakeholders at the very top of the political system when the Prime Minister's office reached out to influencers in 2020 to communicate the government's COVID-19 measures (Pöyry et al., 2022). In Romania, by contrast, influencers were not considered to have a prominent role in political communication until after the presidential elections held in November 2024 were annulled due to suspected foreign interference, including secret payments to TikTok influencers to post political content (Ings, 2025). However, interviews for this study were conducted before these events in Romania.

Our units of analysis are political parties. We focused on established parties, which we define as parties holding seats in the national parliaments. Prior research showed that established parties drive innovation in election campaigns (Koc-Michalska et al., 2016). In addition, we also considered few smaller parties with a distinct digital profile. We conducted qualitative expert interviews (Döringer, 2021) and interviewed political communicators, that is party workers who perform official communicative functions (e.g. media spokespersons, social media managers) and, for example, manage and execute election campaigning. Our plan was to interview at least one such communicator from each established party. We started contacting the parties in each country accordingly by both e-mail and phone calls. In Finland, we were able to interview representatives from the six largest parties. In Germany and Romania, we were not able to recruit interviewees from the right-wing party, despite persistent attempts, but all other established parties are represented in our sample.

All in all, 26 interviews were conducted. Table 1 provides an overview of the participants. All Finnish respondents were interviewed face-to-face, all German respondents via video-conferencing software. For Romanians, we adopted a mixed approach with two face-to-face interviews and seven via video-conferencing software. All interviews were conducted in their native languages (Finnish, German and Romanian), recorded and then transcribed.

We developed an interview guide for organizing the interviews, following the deductive procedure suggested by Kaiser (2014) for expert interviews. In a step-by-step process, we developed the interview questions for the main part of the interviews from the concepts included in the research questions and their dimensions. Appendix A provides a schematic overview of this process and Appendix B includes the interview guide.

We analyzed the data in a qualitative content analysis (Schreier, 2012). To develop the coding frame, we first established a general structure that followed the interview guide. The first author then started developing data-driven subcategories from four of the German interviews, thus creating a first version of the coding frame. This version was passed on to the second author who used the strategy of subsumption to add to the subcategories from four Finnish interviews. Author 1 and 2 then produced a second version of the coding frame during a Zoom meeting. In another loop, this version was delivered to the Romanian authors who implemented the same procedures. This yielded a prefinal coding frame that we pretested on three to four interviews per country. Codings of the pretest were translated into English, distributed among all authors and discussed in a joint meeting a couple of days later. As a result of this meeting, we undertook some minor changes to produce the final coding frame (see Appendix C). In the following main coding, we used this frame to code all of the data, with each author coding the interviews from their country. Codings were translated into English and compiled in an excel document, which served as basis for joint interpretation sessions.

The interviewed political communicators listed several features they wish to see in influencers – many that are commonly considered also in the corporate domain. For example, political communicators look at social media metrics, good fit between party and influencer, as well as reputation-related features. In addition, our interview partners mentioned features that are more unique to the political sphere such as influencers' understanding of political content and dynamics and their ability to explain complex politics to their followers. Another desired feature is a sense of mission in influencers. Political communicators aim at collaborating with influencers who are passionate about a specific social cause or who want to help society in a more general fashion.

This requirement profile produces the ex ante information asymmetries of hidden characteristics and hidden intentions because political communicators need to determine which influencers offer these features. Selection criteria such as social media metrics, fit and qualities relate to hidden characteristics, while the required sense of mission relates to hidden intentions. Political communicators thus need to establish adequate procedures to cope with these asymmetries, and we found that they make do with three approaches.

First, political communicators investigate influencer-created materials, such as posts, channels descriptions and bios. This is what a German interviewee described as “looking deeply into the channel” (Germany, communications team, in other contexts).^[1] However, this approach has its limits, as a Finnish respondent explained: “It's not always clear or easy what their ultimate agenda is. But, of course, many times you can get an idea of the direction when you follow the story a little bit” (Finland, communication team, in election campaigning).

Second, political communicators seek personal contact with influencers. There are various ways of establishing such contact. At an early stage, political communicators try to get to know influencers by interacting with them, be it via emails or via separate chemistry meetings. A Finnish interviewee mentioned that meetings are more informative: “I don't think this is the kind of thing we would start doing just by email, but I believe that there would be some kind of conversation over the phone or face-to-face” (Finland, communication team, no experience). Given the importance of making a good decision about with which influencers to collaborate, meetings are the more risk-averse way of evaluating influencer features. At later stages, when negotiations about possible collaborations become more concrete, political communicators carefully observe how influencers behave during these negotiations to determine their true characteristics and intentions. Another strategy of political communicators is to work only with influencers who are already familiar to them, be it because the influencers are members of their parties or because they belong to the personal networks of communicators. For example, an interview partner (Germany, executive secretary, in election campaigning) strongly argued in favor of this strategy, explaining how the party's regional office recruits influencers solely through and on the basis of personal affiliations with party members who then act as guarantors of the influencers' characteristics and intentions.

Third, political communicators draw on third-party sources to determine influencer characteristics. Common procedures are to conduct Internet searches for information on the influencer and to collect social media metrics. A Romanian communicator pointed to the usefulness of blogs in this matter: “Additionally, we evaluate their online reputation by reviewing articles or blogs written about them” (Romania, head of communications, in other contexts). Another procedure is to ask presumptive experts for their evaluation of influencer features. An obvious way of doing so is working with specialized influencer agencies. A Finnish respondent explained how making use of agency services reduces the risks that come with the process of selecting influencers. The party resorts to agency services, the respondent stated, “to get the right kind of person, because [agencies] have extensive networks” (Finland, communication team, in elections campaigning). However, few interview partners considered this option, mostly because they prioritize other campaign instruments in the face of limited campaign budgets. Interviewees thus showed a certain ingenuity when collecting expert opinions and reported resorting to rather informal procedures such as asking friends and family.

When asked about the difficulty of carrying out reliable evaluations and the effort these evaluations require, answers varied widely. Some interviewees found that influencer selection is easy and can be done with a simple web search and/or a chat with the influencer. Others perceived it as challenging and explained their perception either with their inexperience in working with influencers or the complexity of the task. A German communicator, who belonged to the second group, stressed this complexity by using branding vocabulary: “You must almost carry out a character test to find out whether there really is a brand fit [with the influencer] and brand affinity (…). That's really difficult” (Germany, head of communications, in other contexts). As this statement presages, the individual risk perception played an important role in our interviewees' handling of influencer selection, and they indicated many risk factors that come with influencer collaborations. For example, they were afraid that influencers could spread misinformation or information that is not coherent with what the party communicates. They were also concerned that influencers may be involved in scandals, which then harms the party's reputation. Finally, respondents worried about influencer loyalty. Even committed influencers, they argued, might disapprove of particular party proposals and make their disapproval public. Influencers may also value the loyalty towards their followers higher than the loyalty to the party, as a Finnish communicator explained: influencers “may turn against you when there is criticism from the rest of the community. Like when the social part of social media starts to work in some way. That is why it's different from traditional media” (Finland, head of communications, in election campaigning). This statement underlines how collaborations with influencers are considered to be more risky than traditional media campaigns.

According to our interviewees, influencers can support political parties in attaining two types of campaign objectives. First, influencers can help accomplish self-interested objectives such as reaching new target audiences, cultivating a desired image of candidate and party or highlighting past achievements. These objectives cater to the broader goal of winning votes. Second, collaboration with influencers can foster the common good. This included promoting voter turnout (whether for their party or others) and, more generally, popularizing democracy.

Collaborating with influencers to attain these objectives causes ex post asymmetries of hidden action because political communicators must evaluate whether influencers do the right things to support campaign objectives and whether they devote sufficient effort to this task. Our interviewees described numerous procedures they developed to cope with these asymmetries. These procedures were designed not only to collect information on influencer actions, but also to control influencers to varying degrees. We sorted the procedures into three approaches, depending on their position in the content production process.

First, political communicators resort to procedures that are applied before the beginning of content production. Depending on the type of collaboration, they use contracts and briefings to define their expectations: “We establish clear indicators and set minimum conditions from the beginning. These conditions include specifying the number of posts or stories, the channels to use, and the frequency of posting” (Romania, head of communications, in other contexts). These procedures also aim at determining certain decisions, for example to foreclose that posts are too ridiculous or too scurrilous in tone. Besides, political communicators pursue procedures that grant more leeway to influencers, such as planning posts together or letting influencers take the initiative altogether.

Second, another set of procedures aimed at the stage of content production. Some interviewees reported how they get involved in the production process. However, a Finnish respondent preached caution when doing so: “If you feel you have a creator that you aren't completely sure about, or maybe it’s a difficult topic (…), then we can agree on this kind of a step [i.e. getting involved in content production]. But even that should be done cautiously. Because it easily turns into pre-censorship” (Finland, communication team, in election campaigning). A less intrusive procedure is to obtain information by asking influencers for regular updates on the production process. An interviewee (Germany, executive secretary, in election campaigning) recounted how influencers regularly checked in with the party to make sure the tone they were hitting in a parody video was appropriate. This updating strategy works in the other direction, too. For example, another German informant described how they shared upcoming campaign material with influencers to stimulate related content production. The appeal of this procedure lies in the fact that it can trigger content production in alignment with the campaign without having to explicitly instruct influencers. The informant framed this strategy in terms of “helping out”: “[We messaged:] ‘Here's a campaign video in advance.’ Things like that. Just helping out a bit and providing [influencers] with information so that they have some kind of exclusive materials” (Germany, head of communications, in election campaigning).

Third, our interviewees apply procedures in the phase after content production. Approval processes are an outright way to ensure influencer content meets party expectations. A Romanian respondent made clear that influencers should not publish content without previous approval: “We will request the submission of materials for verification before they are posted” (Romania, communications team, no experience). While this form of hard control was exerted by some of our interviewees, others stressed the need for trust: “I would grant them creative freedom. This is precisely why you selected an influencer – because you've been following them and trust their judgment. After the post goes live, we can examine the metrics to assess its effectiveness” (Romania, communications team, no experience). In this case, information gathering and control are retained by monitoring what happens with influencer content. As a German communication manager highlighted, parties need to “look at the reactions in the community and then deal with them” (Germany, communications team, no experience).

When asked about difficulties in evaluating influencer activities and efforts, answers again varied. In contrast, interviewees widely agreed that it is inevitable to place trust in influencers, at least to a certain degree, because gathering complete information is not possible. When discussing the issue of control, most interviewees echoed the debate on influencer collaborations in the corporate domain that makes a strong case for letting go of control to safeguard influencer authenticity. Beyond this rhetoric of letting go, it was debated more controversially how much control is adequate. We observed that many interviewees felt a certain unease with allowing influencers to operate largely autonomously, as did the Romanian respondent quoted above, who insisted on approval processes. Others linked the need for control to their (un)familiarity with either individual influencers or the instrument in general. The more familiar, the more willing, they said, they are to trust the influencers and let go of control. Control-willing positions faced fierce opposition among some interviewees since controlling influencer conduct “contradicts the nature of influencer collaborations. Nobody [in our team] wants to control them in the sense that they should only say what we want them to say. Essentially, then it's no influencer communication any longer, it's just covert advertising” (Germany, head of communications, in other contexts). This opinion was shared particularly by those respondents who had pursued a career in corporate communication before entering political communication. At the same time, it should be noted that there are many shades of control between the extremes of requiring approval processes for all materials and not interfering with content production altogether.

The previous two subsections presented similarities across countries in outlining common patterns in our data. To further differentiate the findings, this subsection focuses on differences we observed between communication managers in Finland, Germany and Romania. We draw on the country characteristics discussed above to provide some explanatory context to our descriptive observations.

A general difference was that collaborations with influencers were a more established practice in Finland and Germany than in Romania. While most Romanian interviewees emphasized their willingness to work with influencers, they explained they are only just beginning to consider influencer collaborations as part of their campaign strategies. This status is also reflected in the fact that a notable share of interviewees from Romania had not yet worked with influencers. This observation aligns with the country's different history of influencer involvement in politics. The Rezo-case in Germany and the COVID-19 strategy of the Finnish government put influencers on the political agenda early on, whereas Romania witnessed the first broader public debate over influencer involvement into politics only after the annulled presidential elections in late 2024 (and thus after we conducted interviews). Nevertheless, when asked how confident they feel about their abilities to successfully integrate influencers into their campaign, Romanian respondents demonstrated strong confidence: “I've never felt anxious about embracing new social networks. (…) Adaptability is key, and I'm confident in our ability to evolve with the changing landscape of digital media” (Romania, head of communications, in other contexts).

Regarding ex ante asymmetries, we observed that German communicators were more cautious than their Finnish and Romanian counterparts when it comes to weighing the risk factors that come with influencer collaborations. We assume that, again, the history of influencer involvement in politics plays a role here. The first time an influencer appeared on the political stage in Germany was when Rezo harshly criticized the government's climate policies. This incident demonstrated that influencers can not only harm parties, but that they are also difficult to control. In addition, Germans are generally more risk-averse than other Europeans (Reimann et al., 2008), and the Rezo-case may have triggered this aversion. However, this is not to say that German interviewees would not collaborate with influencers, which they do. Rather, the risk factors played a greater role in their considerations. In contrast, Finnish and Romanian communicators were more confident that collaborations would run smoothly.

Regarding ex post asymmetries, we did not identify any specific national patterns in ambitions to control influencer conduct. However, most Finnish and Romanian respondents were comfortable with the idea of paying for influencer services. Meanwhile, German communicators explained that payments would undermine authenticity and thus the persuasiveness of influencer posts. Collaborations without payments are to be preferred, a German communication manager explained, because followers “then have the trust that the person [influencer] is really supporting us and not doing it for money” (Germany, communications team, in other contexts). Another obstacle to influencer compensations in Germany results from the importance that parties attach to influencer marketing. Faced with limited resources in terms of both budget and wo/manpower, German parties tend to prioritize traditional media channels over influencers, while Finnish and Romanian practitioners expressed their willingness to allocate a bigger proportion of their resources to influencers. Shifting the campaign focus (and budgets) to social media makes much sense when viewed in the context of data on social media usage. Social media diffusion rates are significantly higher in Finland and Romania than in Germany (Statista, 2024). Moreover, a higher proportion of Finns and Romanians use social media as a news source, and they also express a stronger preference for political influencer content than Germans do (European Union, 2023).

This study examined how communication managers in political parties from three EU member states – Finland, Germany and Romania – cope with information asymmetries in influencer collaborations. We found that our initial observation that political communicators display a certain hesitancy in regard to strategic influencer communication holds true only to a certain degree. Indeed, many political parties already collaborate with influencers or intend to do so soon. At the same time, these collaborations typically are of a limited scope and nowhere as pivotal as in the corporate sphere.

In investigating possible barriers and constraints that underlie the hesitancy, we identified various procedures political communicators apply to collect information on influencer characteristics and intentions before partnering with influencers (ex ante asymmetries) and on influencer actions during collaborations (ex post asymmetries). We also found that while political communicators are generally confident these procedures help them cope with the asymmetries in an adequate manner, confidence levels varied. Notably, confidence levels depended less on the concrete procedures the communication practitioners had developed, as they used more or less similar procedures across countries. Rather, they depended on individual information needs. In the influencer selection, some political communicators declared themselves satisfied with conducting a web search and/or chatting with influencers, whereas others would have preferred running psychological character test. We explain these differences, at least in part, with the different historical experiences of influencer involvement in politics. During the campaign, political communicators harbor diverging needs for information gathering and, consequently, for controlling influencers, while trying to cater to a rhetoric of letting go of control, as it is dominant in the corporate domain (Borchers, 2023; McMullan et al., 2022). Control ambitions ranged from regularly monitoring influencer actions to granting substantial autonomy to influencers. While we did not observe any notable differences between the three investigated countries in regard to control ambitions, we found that two factors that lie beyond socio-political dimensions play a role here. First, political communicators were more willing to let go of control if they felt familiar with the collaborating influencers or at least with influencer collaborations as a campaign instrument. Second, they displayed a greater willingness to relinquish control if they had a background in corporate communication.

Contracting specialized influencer agencies could help political communicators strengthen confidence. These agencies have built extensive expertise in managing influencer campaigns and established proven procedures for managing information asymmetries. In fact, particularly Finnish communicators tap into agency expertise. However, the costs associated with mandating agencies can be considerable, which brings the issue of limited budgets and the necessity to prioritize back into focus. This cautions communicators to consider the patterns of social media use and the acceptance of political influencer content (European Union, 2023; Statista, 2024) before making budget decisions. Although contracting agencies holds promise, the irony of this approach is that it places political communicators in yet another principal-agent relationship.

This study contributes to previous research on the role of influencers in political communication. While earlier works on this topic have mainly concentrated on either the political content created by influencers (Suuronen et al., 2022; Oliva et al., 2024) or looked at effects on political opinions and behaviors (e.g. Harff and Schmuck, 2023; Naderer, 2023), the perspective of communication managers in political parties is widely under-researched. In adopting this perspective, we found that influencers play an important role in plans for future election campaigns. Communication managers in the three investigated countries regard influencers as powerful allies, even if influencer collaborations can be a risky endeavor. Hence, they are ready to seize the opportunities that arise when influencers engage with politics, such as raising political interest and participation (Dekoninck and Schmuck, 2022; Schmuck et al., 2022).

Communication managers are aware that seizing these opportunities requires skilled craftmanship on their part to avoid the pitfalls the information asymmetries create. For example, they speculate how much control they can expect influencers to put up with. They are also aware of how important it is to select the right influencers. Among others, they display an understanding of the relevance of influencer-message congruence, which experimental research posits (Cheng et al., 2024; Naderer, 2023). Thus, they consider that simply buying influencers' opinions can mean that influencer messages will not unfold the desired persuasive effect. This finding somewhat assuages Woolley's (2022) fear that influencers' involvement in politics threatens democracy. Woolley argues that “many influencers around the globe are spreading political content simply because they are paid to do so” (p. 122), implying that parties could use influencer collaborations to manipulate elections. Yet understanding the need for influencer-message congruence may discourage political communicators from adopting such a strategy.

This study expands research on the management of influencer campaigns in two directions. First, it provides insights into campaign management in the political domain, whereas previous research has focused on corporations (e.g. Borchers and Enke, 2021; Childers et al., 2019). We found that political communicators apply similar influencer selection criteria as their corporate counterparts, such as social media metrics and fit dimensions, but that they also consider criteria that are more specific to the political realm, such as an understanding of political content and dynamics and influencers' general “political-ness.” We also extended the body of knowledge by presenting the first overview of procedures that communication managers use to determine whether influencers meet these criteria. Moreover, this study adds to research on the control of influencer conduct (e.g. Borchers, 2025; Leung et al., 2022; McMullan et al., 2022). We found that political communicators echo the corporate discourse's rhetoric of letting go of control. At the same time, many political communicators do not feel comfortable with letting go. This unease indicates that they have not yet figured out that the “paradigm of letting go” does not require them to truly surrender control, but rather to adopt different, less obvious control procedures (Borchers, 2023). Accordingly, their set of control techniques appears underdeveloped in comparison to the elaborate set that corporate communicators have constructed.

Second, this study suggests a way to theorize influencer collaborations by applying PAT to influencer collaborations. Such applications have been suggested before to overcome the dearth of theorizing client-influencer relationships (Enke and Borchers, 2019; Sundermann and Raabe, 2019), yet, to the best of our knowledge, this is the first study that followed this suggestion. PAT proved to be a valuable tool for investigating the management of influencer campaigns since it allowed a systematic identification of the challenges communication managers face when working with influencers and of the procedures they develop to tackle these challenges.

This study has its limitations. Importantly, the focus on three countries limits the meaningfulness of our data. As reported, we observed many similarities, but also differences between the three countries. Adding further countries to the investigation could possibly reveal that findings we presented as common patterns in this study are essentially specific to our set of European countries. Furthermore, we interviewed a limited number of political communicators from each country and only one communication manager from a right-wing party. Despite this limitation, we argue that our cross-country comparison is adequate because we interviewed political communicators from almost all major parties in Finland, Germany and Romania, thus covering a wide range of political backgrounds and context. In addition, the use of national languages in the interviews and English in the discussions between the researchers carries the risk of details getting lost in translation. Therefore, back-translations were used when necessary to make sure that both the interview questions and responses were understood similarly. Finally, we decided to drop the PAT dimension of hidden information from the interview guide during the early data gathering process as interviews showed that interviewees had difficulties grasping this dimension. This reaction may not come as a surprise, given that the field is only just beginning to explore how to evaluate the campaign contribution of influencers. As a result, however, this study does not provide a complete picture of information asymmetries in party-influencer collaborations.

We want to encourage future research to apply PAT to examine (1) domains beyond the political and (2) the perspective of the agents, namely the influencers. Furthermore, there are other principal-agent relationships in the influencer domain that seem worth investigating. Influencer agencies are an additional actor that enters such constellation, even if respondents in this study hardly made use of them. Followers, too, can be added to the picture as principals who expect to be informed adequately about certain topics in the influencers' niches. These considerations indicate that PAT holds promise to further enhance our understanding of the influencer field.

The supplementary material for this article can be found online.