Actor engagement roles in collaborative settings: a theoretical framework and empirical typology

The concept of actor engagement has been increasingly influential in service research, defining all actors as resource integrators (Storbacka et al., 2016) or resource investors (Brodie et al., 2019). The broad conceptualization of actor engagement enables its application across various contexts and among different types of actors. Nevertheless, this broad perspective focuses on the similarities and does not delineate the different ways actors engage within and across these multiple contexts (Storbacka et al., 2016; Wajid et al., 2019). However, understanding the multitudes of resource investments, dispositions and influence that actors bring to their engagement roles is essential to initiate, maintain and enhance engagement in collaborative settings.

Acknowledging the relevance of a better comprehension of diverse engagement practices and their impact, recent literature explores how engagement (Stadtelmann et al., 2019; Flaig and Ottosson, 2022; Wajid et al., 2023) and co-creation activities (Ekman et al., 2016; Davey and Grönroos, 2019; Klimas and Czakon, 2022) translate into roles. For instance, the investigation of online health communities revealed various engagement practices and, consequently, different actor engagement roles of community members (Stadtelmann et al., 2019). Based on this improved understanding of whether members, for instance, prefer fact-based support or support through the expression of empathy, platform operators can tailor the offerings and opportunities for exchange within the communities.

The academic discussion about actor engagement roles highlights that these roles are dynamic, allowing actors to hold multiple roles simultaneously or shift roles over time. However, there is no unified conceptualization of actor engagement roles and typologies often build on the dichotomy between provider and beneficiary. Such a dyadic perspective fails to capture the complexity of actor constellations in collaborative processes (Storbacka et al., 2016; Jonas et al., 2018). Additionally, research usually remains in the specific setting and the related activities to describe the roles, preventing the application of the results in other contexts (Heikkinen et al., 2007) and tends to focus on system-level roles for organizations, overlooking the role of individual engagement within these collaborations (Jonas et al., 2018). Moreover, the papers only sporadically consider the impact of the various engagement practices of the other actors (Flaig and Ottosson, 2022). However, the mutual influence is particularly important when attempting to manage and ideally enhance engagement in collaborative processes.

Therefore, a better understanding of how resource investment practices can be translated into actor engagement roles in collaborative settings and how these roles induce engagement dynamics within the collaboration is needed. Contributing to such understanding, we seek to address three research questions: (1) How can actor engagement roles be conceptualized in a context-independent manner reflecting actors' investment of resources? (2) Which actor engagement roles can be identified in collaborative settings? (3) How do actor engagement roles in collaborative settings mutually influence each other, inducing engagement dynamics? To answer the research questions, we draw on actor engagement as the domain theory and role theory as the method theory. Contributing to the lack of a unified conceptual understanding in the literature, we first develop a theoretical framework outlining the characteristics of actor engagement roles, namely engagement disposition and engagement activity. We further conceptualize connectedness as the mutual influence among actors that emerges from variations in engagement disposition and activity. Using the framework as a sensitizing device in the empirical investigation, we adopted a qualitative approach and conducted 32 semi-structured interviews with students participating in seminars that required collaboration in small groups. This context enables the exploration of actor engagement roles as they evolve over time, as no pre-established roles or hierarchies, commonly present in organizational settings, exist to influence role formation.

Following the Gioia methodology for data analysis (Gioia et al., 2013), we identified eight actor engagement roles, grouped into three overarching, higher-order categories: resource investment leaders, impeding resource investors and activated resource investors. Moreover, we identified different ways in which the resource investment of the actors holding the various roles influences the resource investment of the other actors, pointing to engagement connectedness and engagement dynamics between the actors in the collaborative setting. Accordingly, we were able to develop an actor engagement role typology and constellation for collaborative settings.

The contribution of our research to the actor engagement literature, and particularly to the actor engagement role literature, is threefold. Firstly, we develop a theoretically derived framework that can be applied across diverse contexts to generate typologies of actor engagement roles. Secondly, by applying this framework in a collaborative setting without pre-existing roles, hierarchies or monetary exchange, we develop a typology and constellation of actor engagement roles within this specific context. Thirdly, we advance knowledge of connectedness by strengthening its theoretical foundation through clarifying that connectedness is not a defining characteristic of actor engagement roles but rather their emergent consequence, and by empirically illustrating actor-induced engagement dynamics in collaborative settings. In practice, the actor engagement role framework is valuable for managers or facilitators, for instance in intra- or inter-organizational projects, as it enables a nuanced understanding of individual team members' engagement and supports sustaining engagement across partners. The typology and constellation may, for instance, support role allocation and engagement management in newly formed groups, such as emerging movements and NGOs.

The structure of the paper is as follows. Firstly, we provide an overview of the theoretical background and current literature about actor engagement and role theory. Building on the theoretical foundation of actor engagement, we generate an actor engagement role framework by outlining the defining characteristics. The following section describes this paper's research approach, context, design and data analysis process, followed by the presentation of the results. We conclude with a discussion of the findings, implications for academics and practitioners, acknowledging the limitations of this research and highlighting avenues for future research.

Actor engagement is a relatively young concept originating from the field of marketing and is closely linked to the theoretical perspective of Service-Dominant (S-D) Logic. S-D Logic provides a meta-theoretical framework for understanding value co-creation and value co-destruction processes among interacting actors (Storbacka et al., 2016; Wajid et al., 2019; Lumivalo et al., 2024). Actors are conceptualized as resource integrators (Storbacka et al., 2016) or resource investors (Brodie et al., 2019). This understanding is captured in the actor engagement concept, which is considered a mid-range concept within S-D Logic (Alexander et al., 2018). Accordingly, the perspectives of value co-creation and value co-destruction differ from the actor engagement perspective in that their primary focus lies on value outcomes, namely value formation or value decline, whereas actor engagement focuses on the micro-foundation of resource investment itself. Consequently, actor engagement can be understood as a mid-range concept that is orthogonal to the perspectives of value co-creation and value co-destruction. It is thus an observable and complementary concept that contributes to a deeper understanding of value co-creation and value co-destruction processes in service systems (Storbacka et al., 2016; Lumivalo et al., 2024).

Specifically, actor engagement is defined as “both the actor's disposition to engage and the activity of engaging in an interactive process of resource integration within the institutional context provided by a service ecosystem” (Storbacka et al., 2016, p. 3009). In line with the definition, actor engagement encompasses all actors in an ecosystem, including employees, suppliers, distributors, government, the media and other publics (Kleinaltenkamp et al., 2019). Thus, it broadens the scope of engagement, as it no longer differentiates between actors and can be applied in various contexts to examine engagement, for example, in the higher education context (Sim et al., 2018). In addition to its generic nature, the concept of actor engagement differs from concepts such as customer or employee engagement in that it does not adopt a dyadic perspective. Instead, it considers not only the engagement between an engagement subject and an engagement object, such as between customer and provider or employee and work, but also the influence of other actors and the institutional context (Brodie et al., 2019).

Actors can be understood as either individual humans or collections of humans, such as organizations, who are usually categorized according to their discrete roles and functions (Lusch and Vargo, 2014). To get a more fine-grained view of the actors and their resource investment, and thus their roles and functions, the engagement literature has started to combine actor engagement and role theory (e.g. Flaig and Ottosson, 2022).

Role theory examines characteristic patterns of human behaviour (Biddle, 1986). Traditional role theory conceptualizes roles as “comprehensive pattern[s] for behaviour and attitudes” (Turner, 1979, p. 124), highlighting both behavioural and cognitive elements as integral components of roles (Lynch, 2007). The activities and attitudes that make up a role are recognised and shared within a group or a social system (Wittmayer et al., 2017). An actor can hold several roles, and this set of roles defines the social position of the actor within a social structure and, thus, in a social system (Edvardsson et al., 2011).

Several papers on the topic of actor roles apply the perspective of S-D logic (Ekman et al., 2016; Stadtelmann et al., 2019; Wajid et al., 2023) or refer to the perspective of S-D logic (Davey and Grönroos, 2019; Ekman et al., 2021; Flaig and Ottosson, 2022). Within this theoretical perspective, a social role is defined as “a set of practices (repeated activities) that connect one actor to one or more other actors” (Akaka and Chandler, 2011, p. 251). We apply the definition of a social role (Akaka and Chandler, 2011), transferring it to the field of actor engagement by defining a role as a set of engagement practices that connect one actor to one or more other actors.

The connection to other actors points to the concept of role constellations defined as “networks of roles that interact with each other through certain activities” (Maria et al., 2021, p. 49) or as “webs of roles, which interact, interrelate and co-evolve with one another” (Wittmayer et al., 2017, p. 50). Focusing on the individual role and the role constellation highlights the relationship between the roles (Wittmayer et al., 2017). Considering collaborative settings comprising multiple actor engagement roles, it is crucial not only to develop a typology of actor engagement roles, highlighting the different engagement practices of each role but also to consider a role constellation, since roles co-evolve with and influence one another based on actors' engagement within the collaborative process.

Some researchers have begun to create typologies of actor roles, basing their typology on the engagement of the actors (Stadtelmann et al., 2019; Flaig and Ottosson, 2022; Wajid et al., 2023) (Table 1) or on co-creation activities (Ekman et al., 2016; Davey and Grönroos, 2019; Klimas and Czakon, 2022) (Table 2). The papers referring to co-creation activities apply the perspective that actor engagement presents the micro-foundation for co-creation (Storbacka et al., 2016). Since the focus of the studies is primarily on value co-creation, no negatively connoted roles or negative influences of roles are presented, except in the study by Flaig and Ottosson (2022), which examines market-shaping roles and identifies market rivals and market detractors. All papers rely on different characteristics to identify the various roles. However, most of the papers characterise the roles based on the observable activities of the actor (e.g. Stadtelmann et al., 2019; Klimas and Czakon, 2022), whereas a few papers have already started to go beyond what is observable to describe the roles by pointing to the actor's readiness (Davey and Grönroos, 2019; Wajid et al., 2023) or disposition (Flaig and Ottosson, 2022). Flaig and Ottosson (2022) are the sole researchers who employ connectedness as a defining characteristic in the conceptualization of roles. Most typologies, except Ekman et al. (2016) and Klimas and Czakon (2022), are based on context-specific activities of the actors within the collaboration (e.g. Davey and Grönroos, 2019; Stadtelmann et al., 2019; Wajid et al., 2023). Although all papers adopt a collaborative perspective, considering actors in ecosystems, systems, networks, markets or platforms, the provider/beneficiary dichotomy is still frequently referenced (Ekman et al., 2016; Davey and Grönroos, 2019; Stadtelmann et al., 2019; Wajid et al., 2023). Moreover, the perspective adopted in most studies focuses primarily on actors at the institutional level. The only exceptions are typologies from the healthcare domain, which consider both healthcare providers (institutional actors) and patients (individual actors) (Davey and Grönroos, 2019; Stadtelmann et al., 2019).

Given the inconsistent understanding and conceptualization of actor (engagement) roles in the literature, this paper first aims to develop an actor engagement role framework. Drawing on the actor engagement concept, we outline the key characteristics that define an actor engagement role. Due to the generic nature of the concept, the resulting framework can be applied to identify actor engagement roles across diverse contexts, encompassing both individual and collective actors. We apply this framework to our empirical data to develop an actor engagement typology and constellation for collaborations, particular collaborative settings without pre-existing roles, hierarchies and monetary exchange. Unlike prior research focusing on system-level roles, our typology addresses the individual level. Moreover, we develop a constellation that illustrates how roles co-evolve and interact, conceptualizing engagement connectedness as emerging from these roles and highlighting engagement dynamics within collaborative settings driven by the diverse engagement practices of different roles. Accordingly, our typology and constellation extend beyond the frequently referenced provider/beneficiary dichotomy and account for the complexity of collaborative settings.

Actor engagement provides an appropriate theoretical foundation for examining role differentiation in collaborative settings because it conceptualises engagement as the voluntary investment of resources in interaction with other actors within a service system (Storbacka et al., 2016; Brodie et al., 2019). As a mid-range concept grounded in S-D logic, it offers an analytically tractable way to examine actor-level processes while remaining consistent with broader value co-creation and value co-destruction perspectives and extending these by explicitly focusing on the micro-foundations of actors' resource investments (Vargo and Lusch, 2011; Alexander et al., 2018; Lumivalo et al., 2024). Recent actor engagement scholarship has increasingly emphasised engagement as a resource investment process embedded within interconnected service systems and characterised by variations in engagement disposition, engagement valence and actor connectedness (Brodie et al., 2019; Hollebeek et al., 2022). At the same time, actor engagement research is increasingly concerned with understanding how engagement contributes to broader collaborative, societal and ecosystem-level outcomes across domains such as healthcare, education, public services and sustainability (Alexander et al., 2025; Matos et al., 2025; Muñiz-Martínez, 2025). However, comparatively less attention has been directed towards understanding how different configurations of engagement manifest at the actor level and how these configurations shape engagement dynamics within collaborative settings. We therefore draw on these developments to integrate engagement disposition and the activity of engaging into a unified framework for conceptualising actor engagement roles, while understanding connectedness as an outcome of variations in engagement disposition and activity.

Building on the theoretical foundations of the actor engagement concept, we outline the defining characteristics of actor engagement roles (Figure 1). Different role types can subsequently be identified based on the configuration of these characteristics. The two most frequently cited definitions of actor engagement are as follows: according to Storbacka et al. (2016, p. 3009), actor engagement encompasses “both the actor's disposition to engage and the activity of engaging in an interactive process of resource integration within the institutional context provided by a service ecosystem”, while Brodie et al. (2019, p. 2) define actor engagement as “a dynamic and interactive process that reflects actors' disposition to invest resources in their interaction with other connected actors in a service system”. In the second definition by Brodie et al. (2019), the disposition of the actor is emphasized, while the actual resource investment is only implied. In contrast, the first definition by Storbacka et al. (2016) explicitly outlines that actor engagement encompasses both the disposition and the activity of engaging. Accordingly, actor engagement roles should be defined both by their engagement disposition and the activity of engaging.

Engagement disposition can be understood as the readiness to invest resources (Brodie et al., 2019) and comprises capacity and tendency (Sim et al., 2022). Tendency refers to the willingness of the actor to invest resources, while capacity refers to the resources available to the actor (Sim et al., 2022). The engagement disposition can vary on a spectrum from positive to negative (Li et al., 2018), indicating its valence and thus the positivity or negativity of the readiness to invest resources, which reflects positive or negative feelings and an affirmative or dissenting evaluation of the engagement object (Grégoire and Fisher, 2008; Jahn and Kunz, 2012). In the literature, engagement disposition is frequently conceptualized as preceding the activity of engaging, with the assumption that the valence of the disposition also determines the valence of the engaging activity (Wang et al., 2013). However, a reverse effect from the activity of engaging on the disposition has also been observed (Fehrer et al., 2018); thus, a mutual influence between both components can be assumed.

As the disposition describes the readiness to invest resources (Brodie et al., 2019), the activity of engaging refers to the actual investment or integration of resources. The activity of engaging is defined by its intensity. Intensity can be understood in relation to the amount of invested resources and the duration of the resource investment (Wang et al., 2023). If the same amount of resources is invested within a short period of time, the intensity is higher than if the resource investment is spread over a longer period. Moreover, the activity of engaging can differ regarding the various types of resources, pointing to the different dimensions of engagement. Usually, resources are differentiated into cognitive (thought-based), emotional (feeling-based) and behavioural resources (activity-based) (Hollebeek and Andreassen, 2018). We can further differentiate between different forms of resources within these multiple dimensions. For instance, the investment of behavioural resources (behavioural dimension) in a social media context includes both forms of resource investment: viewing and sharing (Van Doorn et al., 2010). The different forms of resources can vary between more and less active, as viewing is less active than sharing (Wang et al., 2023). Just like the engagement disposition, the activity of engaging can also manifest along a spectrum from positive to negative, reflecting a corresponding valence (Li et al., 2018).

The concept of actor engagement understands resource investment in “interaction with other connected actors in a service system” (Brodie et al., 2019, p. 2). Thus, engagement does not exist in isolation but depends on the context and the other actors within the systems surrounding the actor, pointing to the importance of connectedness as a critical property of actor engagement (Brodie et al., 2019). Accordingly, the resource investment of the focal actor can influence, or be influenced by, the resource investment of other actors (Fehrer et al., 2018). When these mutual influences lead to changes in engagement, understood as changes in the investment of resources, we refer to this as actor-induced engagement dynamics. Hence, we do not consider connectedness as a characteristic of actor engagement roles but rather conceptualize it as an emergent consequence of the varying manifestations of engagement disposition and the activity of engaging. As actor engagement always occurs through interaction with other actors, mutual influence is inherent to the concept. This perspective guides our conceptualization throughout the framework, the empirical typology and constellation and the conclusions and implications.

Given the limited prior research and the absence of a unified conceptualization of actor engagement roles, the first step of this research was to theoretically derive an actor engagement role framework. To further examine emerging actor engagement roles and actor-induced engagement dynamics in collaborative settings, we followed a qualitative, inductive-informed approach, using the framework as a sensitizing device rather than as a validation structure. Building on the definitions of actor engagement by Storbacka et al. (2016) and Brodie et al. (2019), we understand collaborative settings as contexts characterized by interactions between connected actors oriented towards a shared engagement object in a service system. In line with S-D logic, such collaborative settings can be conceptualised as service systems, as actors integrate and exchange resources to achieve mutually defined goals, thereby enacting service-for-service exchange within a shared institutional context (Vargo and Lusch, 2011). In addition, there must be a shared engagement object, meaning a common goal of the collaboration. Since actor engagement roles reflect roles and positions within social structures that are defined solely by actors' engagement rather than by pre-existing positions such as professional roles, hierarchical status, skills, expertise or prior relationships, this phenomenon is best examined in a context characterized by the following conditions: a setting in which groups form entirely anew and collaborate over a defined period toward a shared goal. Ideally, the actors should not know each other beforehand and therefore have no prior knowledge of the personalities or expertise of the other actors. Furthermore, groups should be externally assigned so that participants cannot choose their collaboration partners. Such conditions are difficult to observe in established organizational settings, where prior relational histories shape behavioural expectations and may obscure engagement-based role emergence. Following an evaluation of alternative empirical settings, the student collaboration context was therefore identified as theoretically appropriate, as it allows examination of how engagement configurations alone give rise to role differentiation and was employed to address research questions two and three. However, it should be noted that the ideal conditions for the formation of actor engagement roles rarely occur in their pure form, both in other contexts and in the context selected for this study. Nevertheless, we deliberately chose a context that most closely approximates these conditions. Consequently, when transferring the findings of this study to other contexts, the specific contextual conditions and their corresponding influence on role formation must be carefully considered.

The study involved university students from three courses at a German university, where they collaborated in small, randomly assigned groups to fulfil course requirements. These courses included bachelor and master programs with a diverse mix of national and international participants, lasting between seven weeks and an entire semester. In total, 32 semi-structured interviews were conducted, which represented 45% of the eligible students. The average length of the interviews was 37 min (minimum of 24 min and maximum of 61 min). In total, 19 h of audio material were generated. Interviews were conducted in English via Zoom by one researcher to ensure consistency. The data collection aimed for theoretical saturation, employing Shah and Corley's (2006) constant comparison method to extend interviews until no new information or themes emerged. A semi-structured interview guideline (see Web Appendix for the full guideline) was used to cover essential concepts while allowing participants to introduce additional topics. Questions explored personal engagement, perceptions of team members' engagement and the development of engagement over time. Accordingly, respondents were asked about dynamics within bounded timeframes (between seven weeks and an entire semester), which renders a single-point-in-time data collection appropriate, as the study does not investigate longitudinal evolution. Participants also discussed their perceived engagement roles and reciprocal influences between their engagement and their team members' engagement. Accordingly, it should be noted that the interviewees were explicitly asked about their perceptions and therefore reflect a subjective perspective. To address this, we triangulated the interviewees' statements within each group to identify and account for divergent perceptions. While activities such as task allocation could be comparatively well aligned through this triangulation, perceptions of other actors' capacities and capabilities may nevertheless diverge across respondents. However, given that roles constitute socially constructed phenomena and their influence on focal actors is particularly contingent on subjective perception (Wittmayer et al., 2017), capturing these perceptions remains especially relevant for the present analysis.

The interviews were audio-recorded and transcribed into 255 A4 pages using Trint transcription software. All interviews were conducted and transcribed in English, although English was not the interviewees' native language. However, all courses were taught in English; therefore, it was assumed that the interviewees possessed a sufficient level of English proficiency to participate in the interviews. During transcription, no editing was applied. After familiarization with the data, we followed the Gioia methodology (Gioia et al., 2013) for a systematic analysis using the software MAXQDA. The Gioia methodology assumes that the participants are knowledgeable agents whose terms and codes are utilised in a first-order analysis of the data. In the second-order analysis, researcher-centric themes and aggregated dimensions are considered (Gioia et al., 2013). Accordingly, first-order codes were initially derived inductively from the interview data. During the second-order analysis, the first-order codes were assigned to the characteristics of the actor engagement role framework, which was used as a sensitizing device, thereby following an inductive-informed approach. Then patterns in these characteristics were consolidated into aggregated dimensions and designated as specific roles. All authors were involved in the various iterations, going back and forth between transcripts and the coding and the related generation of roles and their constellation, discussing the insights emerging from the data and the theoretical background in a series of feedback sessions. An independent coder reviewed and recoded the data to ensure intercoder reliability, following DeCuir-Gunby et al.'s (2011) approach. Intercoder reliability was computed at 0.82, surpassing the minimum level of 0.7 intercoder reliability (Miles and Huberman, 1994; Rust and Cooil, 1994).

During the data analysis, it became apparent that there are three major groups of roles. Firstly, the interviewees emphasised the leader roles and designated them directly as leaders or heads of the collaborative process. They were consistently identified as essential for the collaboration, typically one to two actors per group assuming these roles. Secondly, there were usually one to two actors who held roles perceived as not contributing or hindering the collaboration. Thirdly, the majority of roles were categorised as middle ground. These roles are diverse, but none stood out as distinctly positive or negative, unlike the more extreme roles described earlier. All roles of this middle category had in common that their engagement needed to be activated by other actors. Based on this rough categorization, which was clearly discernible in the data and mainly derived from the perspective of the respondents regarding the influence of roles within the collaborative process (clearly positive, negative or neither particularly positive nor negative), we generated the categories of the leading resource investors, the impeding resource investors and the activated resource investors (Figure 2). We then delved deeper into the data to examine in detail what characteristics distinguish these role categories, and which roles can be identified within these categories. For this purpose, we applied the theoretically derived role framework as a sensitizing device and identified the following roles across the three categories. Within each category, roles are differentiated according to the actor's engagement disposition, reflected in varying degrees and valences of both tendency and capacity to invest resources and the activity of engaging, reflected in varying levels of intensity, dimensional expression and valence. The role names introduced below reflect these configurations, with each label capturing a distinctive pattern of resource investment as it emerged from the data.

The high resource investor role and the accelerating resource investor role belong to the category of leading resource investors. Both roles are attributed a leadership position, with leadership in the case of the high resource investor being primarily content-oriented, whereas leadership in the case of the accelerating resource investor is more process-oriented. The category of impeding resource investors comprises the scarce resource investor role and the denialist resource investor role. The roles differ in that the scarce resource investor possesses no or only limited resources, whereas the denialist resource investor exhibits a high resource capacity; however, both are unwilling to invest resources or may even invest negative resources. The largest category, activated resource investors, includes the temporary high resource investor role, the supporting resource investor role, the deprived resource investor role and the reluctant resource investor role. While very diverse in nature, actors holding these roles are all characterized by the fact that they did not become active on their own but were activated by other actors, who often assume a leading resource investor role. Additionally, these roles exhibited similar characteristics to the more extreme roles in a mitigated form, meaning that similar engagement practices are demonstrated by, for instance, the accelerating resource investor role (leading resource investor category) and the supporting resource investor role (activated resource investor category) but with varying degrees of manifestation. Thus, we arrived at the depiction of the roles illustrated in Figure 2, where the activated actor roles are positioned in the inner box, demonstrating similar tendencies and capacities of resources as well as similarities in actual resource investment to the more extreme roles (presented in the outer box), albeit in a milder form.

The high resource investor role belongs to the overarching category of leading resource investors. It is defined by a high tendency and capacity to invest resources, reflecting a positively valenced engagement disposition. The role is characterised by the investment of a substantial amount of emotional, cognitive and behavioural resources. These invested resources are content-related, show a high activeness and are invested during the entire collaborative process.

Actors assuming this role possess a range of resources, such as knowledge, skills and experience. In the words of one participant who described the engagement of another team member having the role of the high resource investor:

The quote shows the high capacity of resources in the form of previous experience and the ability to create a theoretical model based on the data. The last sentence also highlights their willingness to invest their experience and skills by offering to build the model independently, thus showing a high tendency.

The actual investment of a substantial amount of cognitive and behavioural resources that are content-related manifests, for example, in the high workload taken over by actors holding this role:

The description highlights that the high resource investment drove the collaboration content-wise and illustrates particularly the investment of behavioural and cognitive resources, which was observable by the other actors. The crucial amount of invested emotional resources is expressed, for instance, in the excitement about the project illustrated in the following quote by a participant holding the high resource investment role:

When asking participants about the influence of the high resource investment of actors holding this role, they indicated that this engagement practice can have both a reinforcing and a diminishing effect. The high resource investment can decrease the resource investment of the other actors since they rely on the high resource investor and perceive no necessity for additional effort on their behalf, as reflected in this quote:

The quote highlights that the combination of high resource capacity, in the form of good writing skills and a high tendency to invest these resources, demonstrated by the actor's enjoyment of working on the paper, led the participant quoted here to assume that the outcome would be good even without their resource investment. On the other hand, the high resource investment associated with this role can also increase the resource investment of the other actors. This is because the other actors are motivated to invest more resources, for instance, to avoid unfavourable comparisons and meet the corresponding expectations, since in collaborations, it is often assumed that approximately the same amount of resources should be invested by each actor:

Actors assuming the accelerating resource investor role, and thus the second role of the resource investment leader category, take on an activating and organising function, distributing tasks and creating structures so others are enabled to invest their resources. Thus, they invest a high amount of process-related emotional, cognitive and behavioural resources characterised by high activeness during the entire collaboration. This role holds a positively valenced disposition, comprising a high tendency to invest resources but a low capacity of content-related resources.

The collaboration outcome is paramount for actors assuming the role of the accelerating resource investor. Hence, they exhibit a high tendency to invest resources in leading the whole team to the intended results:

The quote shows the high tendency described as “high ambitions or aspirations” (T19) and demonstrates the actual activity of engaging in this role. The resource investment of this role focuses on activating and directing the resource investment of other actors within the collaboration. This is done with the ambition to ensure that all actors collaboratively achieve good results. Since the resource investment associated with this role primarily pertains to the engagement of other actors, a high resource capacity in the form of content-related knowledge and experience is not required:

The quote illustrates that the actor took over the role of the accelerating resource investor since she had less experience than the other actor, who performed the role of the high resource investor. The investment of cognitive and behavioural resources can be illustrated, for instance, by the generation and allocation of tasks or by reminding others about tasks or meetings, as illustrated in the quote above. But actors holding this role also contribute emotional resources, such as motivating other actors and making jokes to evoke a positive collaboration atmosphere:

This quote also illustrates that the resource investment in this role does not pertain to the content development of the collaborative process but consistently focuses on the other actors and their engagement. The investment of emotional, cognitive and behavioural resources of this role activates the resource investment of the different actors but also encourages and guides the ongoing resource investment by, for example, generating and distributing tasks and motivating the other actors:

All three quotes show that the form of process related resources invested is crucial for this role and its influence on the other actors.

The scarce resource investor role belongs to the category of impeding resource investors. It is defined by low tendency and low capacity for resource investment, reflecting a negatively valenced engagement disposition. The role is characterized by the investment of no resources or resources characterized by negativity, demonstrating low activeness throughout the entire collaboration.

The actors holding the scarce resource investor role possess very few resources they could invest, and they are unwilling to invest the available resources. This is expressed, for example, in the student context of lacking language skills and the associated unwillingness to contribute resources to the collaborative process:

The quote shows the interrelation between low resource capacity or perceived resource capacity, such as language skills and the following low tendency of insecurity to invest resources in the collaborative process. However, this actor is convinced that they will invest the necessary resources in the future once they have access to them. A contrasting attitude is demonstrated by another actor holding the scarce resource investor role:

The absence of active participation and the early departure from scheduled meetings illustrate the low tendency to invest resources and display the actual investment of no resources. The investment of resources characterized by the negativity of the scarce-resource investor is presented in the following quote:

This quote exhibits the investment of negatively valenced resources by complaining without offering alternative ideas and the influence of this resource investment practice on the other actors. The participant providing this quote mentioned that this behaviour made them mad. This indicates a shift from positively valenced emotional engagement to negatively valenced emotional engagement or the increase in negatively valenced emotional engagement induced by the engagement practice of an actor holding the scarce resource investor role. In addition to this influence on the activity of engaging, the role of the scarce resource investor can also affect the engagement disposition of the other actors:

The feeling of demotivation shows a negative impact on the tendency to invest resources. However, in this case, resources are still invested because the actor cares about the collaborative process and its success and feels that the lack of resources must be compensated. As a result, the investment in behavioural resources even increases. However, the resource practice of the scarce resource investor can significantly impact the tendency of the other roles to invest resources, leading to a decrease in their activity of engaging. This is evident in the following quote from an actor, who perceived themselves as the leader of the collaboration in conjunction with another actor:

The second role in the impeding resource investor category is the denialist resource investor role. This role is defined by low tendency but high capacity for resource investment, reflecting a negatively valenced engagement disposition. A characteristic of this role is the investment of no resources or resources marked by negativity, demonstrating low activeness throughout the entire collaboration process.

The denialist resource investor role holds resources, but these resources are not being invested because actors having this role are unwilling to do so. There are different reasons why the tendency to invest resources is low in the student context presented, for instance, due to a lack of time because of multiple simultaneous deadlines and submissions or prioritization of other activities, such as part-time jobs, internships or family circumstances, as described, for example, in the following quote:

The actual resource investment of the denialist resource investor role manifests in the same way as the engagement of the scarce resource investor by investing no resources (or only a negligible amount of resources) or resources characterised by negativity. One example of the lack of resource investment of the denialist resource investor is the following quote, which describes that an actor holding this role did not participate in partner work within the collaboration:

While the quote does not indicate why the actor did not invest resources, i.e. contribute to the presentation, it can be inferred from the entire interview that the actor indeed had resources but did not invest them due to a lack of time caused by a concurrent internship.

Since the engagement of the two roles, reflective of the impeding resource investor category, is expressed in the same way – by the investment of no resources or resources characterised by negativity – the influence on the other actors is also identical:

It can transform previously positively valenced emotional engagement into negatively valenced emotional engagement or increase the negatively valenced emotional engagement. Additionally, it can reduce the tendency to invest resources, which may not necessarily lead to a decrease in actual resource investment, but it can. Conversely, it might even result in an increased investment of behavioural resources to compensate for a lack of resources. The last case is illustrated in the following quote, as one student describes the resultant lack of engagement of another student, leading to frustration but at the same time increasing their behavioural resource investment:

One role belonging to the activated resource investor category is the temporary high resource investor role. This role is defined by the temporary high tendency and capacity for resource investment, reflecting a positively valenced engagement disposition. The temporary investment of a substantial amount of emotional, cognitive and behavioural resources characterises the role, with the invested resources being content related and showing a high level of activeness.

The temporary high tendency, high capacity and high investment of resources are related to a specific task of the collaborative process. For instance, students were responsible for distinct activities like conducting interviews, creating a persona or designing the project. Once these tasks were allocated to the students, they felt accountable. Since the skills of the students usually determined the tasks they would be allocated, students taking over this role have the capacity and the tendency to invest the resources for the specific time they are working on the particular task and thus invest a substantial amount of emotional, cognitive and behavioural content-related resources:

Other actors usually activate the resource investment of the temporary high resource investor role, for instance, in a discussion about task allocation, as indicated by the following quote of an actor holding the accelerating resource investor role:

By comparing the temporary high resource investor role and the high resource investor role, one can observe the similarity in their engagement practices, in varying degrees of intensity. Actors with the temporary high resource investor role are also occasionally perceived as leaders, albeit not for the entire collaboration but rather for a specific task within the context of which they exhibit an extremely high resource investment.

The supporting resource investor role is also part of the activated resource investor category. It is defined by high tendency and low capacity, reflecting a positively valenced engagement disposition. The role is characterised by the investment of an average amount of behavioural, emotional and cognitive resources. These resources are process-related and can be categorized within the medium range of activeness.

Actors holding this role do not possess resources or only a few resources. However, they are willing to contribute to the collaborative process and thus invest their limited amount of resources:

Because actors holding this role have a high tendency but lack capacity, they are primarily assigned tasks that do not require extensive resources and are process related, such as administrative tasks of formatting documents, creating folders, doing the layout for presentations or recruiting interview partners.

The investment of the resources is activated by the other actors, usually by directly allocating specific tasks. This was illustrated in the first quote when the actor expressed that they “allocated the most easiest tasks” (T30) to the actor holding the supporting resource investor role and is also highlighted in the following quote:

When the actors in this role take over more organizational tasks, such as creating folders and setting up Zoom meetings, this type of resource investment shows similarities to the activating and guiding function of the accelerating resource investor role but on a lower level since these tasks are less crucial for the overall process.

The deprived resource investor role is also assigned to the activated resource investor category. This role is defined by low tendency and low capacity of resource investment, reflecting a negatively valenced engagement disposition. The role is characterized by the investment of a low amount of behavioural, emotional and cognitive resources categorized within the medium range of activeness.

The low tendency and low capacity of resources of this role result in a lack of quality if the actors holding this role are doing the tasks that are assigned to them:

The mentioned low expectations regarding work quality illustrate the deprived resource investor role's low tendency. The missing information in the elaborated text indicates that this role lacks the appropriate resources to implement tasks adequately. The low amount of resources that actors invest in this role, and in this example particularly the low amount of cognitive resources, are also described by one participant as follows:

Since the tendency and capacity of actors holding this role are low, the activation of the resource investment must be more intense than for the other roles of the activated resource investor category with high tendency:

The resource-deprived and scarce-resource investor roles are similar regarding their engagement disposition. Still, they differ in the sense that actors holding the scarce resource investor role are investing no or negative resources. Actors holding the deprived resource investor role invest a few resources, thus contributing to the collaboration but on a shallow level and unwillingly.

The role of the reluctant resource investor is also accounted for among the activated resource investor category. The role is defined by low tendency but high capacity of resource investment, reflecting a negatively valenced engagement disposition. The role is characterized by the investment of a low amount of behavioural, emotional and cognitive resources categorized within the medium range of activeness.

The low tendency but high capacity of this role can be reflected in an attitude of not caring about the collaboration and thus showing a lack of willingness to invest resources:

The low tendency to invest resources, in turn, leads to a low amount of behavioural, emotional and cognitive resources, which is expressed in the missing effort allocated to the work and the necessity to remember the actors holding this role to invest resources:

To activate the resource investment of this role, actors must be directly addressed, reminded and encouraged:

The reluctant resource investor role and the denialist resource investor role differ in the extent of the tendency and capacity to invest resources. Actors holding the denialist resource investor role often possess many resources but are unwilling to invest them because of a lack of interest or time. The actors holding the reluctant resource investor role have fewer resources than the reluctant resource investors, and even if they are not willing to invest them proactively, they are investing them if required. In contrast to the denialist resource investor role, whose lacking or negative resource investment does not contribute to the collaboration, the resource investment of the reluctant resource investor does on a low level.

Our study highlights how actors' engagement determines their position and function in collaborations characterised by the absence of pre-existing hierarchies, roles and monetary exchange. The developed typology and constellation (Tables 3–5) describe actor engagement roles based on the engagement disposition and the activity of engaging and show how the different engagement practices influence the engagement of the other actors (connectedness). Our findings emphasize the importance of considering both engagement activity (e.g. Ekman et al., 2016; Klimas and Czakon, 2022; Stadtelmann et al., 2019) and disposition (Davey and Grönroos, 2019; Flaig and Ottosson, 2022; Wajid et al., 2023). This aligns with role theory, defining roles as comprehensive patterns of behaviour and attitudes (Turner, 1979).

In comparing our typology with existing literature, similarities emerge with the work of Flaig and Ottosson (2022) and Klimas and Czakon (2022). Firstly, drawing from the activities of actors in co-creation processes within innovation ecosystems, Klimas and Czakon (2022) identified four roles in co-creation processes: direct value creation, supporting value creation, encouraging entrepreneurship and leadership. Direct value creation aligns with our high-resource investor role, focusing on content-wise contributions. Klimas and Czakon (2022) describe the leadership role pointing to coordination and communication, which corresponds with our resource investor accelerator role. The supporting value creation role presents similarities with the activated resource investor roles. These parallels display that in co-creation processes on an institutional level and collaborative processes on an individual level, roles of the same type emerge. This could be attributed to the fact that even in collaborations between institutions to generate innovations, individuals from the organizations collaborate (Alexander et al., 2018). Our typology extends beyond Klimas and Czakon (2022) by providing a detailed description of roles organized into three overarching categories. Existing co-creation role frameworks tend to classify actors by how they contribute to value creation. In contrast, our typology differentiates actors based on how they invest resources and considers their disposition, intensity, dimensions and valence. This allows us to capture negatively valenced roles and situations in which actors contribute behaviourally while simultaneously undermining others'emotional engagement, dynamics that are not typically foregrounded in co-creation role classifications. Our typology highlights important distinctions among roles, suggesting varied activation requirements for individual roles.

Secondly, when acknowledging that actor engagement always occurs in connection with other actors (Brodie et al., 2019), it becomes imperative to consider the influences between the engagement practices of the actors. In the current literature, this is done only by Flaig and Ottosson (2022) in a related yet different field of research, who are investigating specifically market-shaping roles based on market-shaping engagement. Even though Flaig and Ottosson define connectedness as a characteristic of the role, and we define roles only through specific engagement disposition and activity of engaging, and the influence on the engagement of other actors as a result of the specific engagement practices, both studies highlight the importance of considering engagement connectedness. Flaig and Ottosson (2022) presented the influence in a rather abstract manner. For example, they described the engagement connectedness of the market driver roles as “impacts on other market actors negatively” and “influence behaviour of other market actors” (p. 73). Our results provide more nuanced insights, for instance, by showing that only the behavioural engagement of an actor can be increased, while simultaneously converting positively valenced emotional engagement into negatively valenced emotional engagement through the engagement practices of actors holding one of the impeding resource investor roles.

Flaig and Ottosson (2022) stand out as the sole authors who explicitly incorporate connectedness into their analysis and, at the same time, uniquely account for potentially negatively associated roles and their adverse effects by including categories such as market rivals and market detractors. However, their conceptualization of connectedness, particularly with regard to the market detractor, differs substantially from the influence exerted by impeding resource investors. Whereas the market detractor role is characterized as having no influence on other market actors, impeding resource investor roles increase other actors' activity of engaging, while simultaneously exerting a negative effect on their disposition and emotional engagement.

This research was guided by three research questions: (1) How can actor engagement roles be conceptualized in a context-independent manner reflecting actors' investment of resources? (2) Which actor engagement roles can be identified in collaborative settings? (3) How do actor engagement roles in collaborative settings mutually influence each other, inducing engagement dynamics? The results emerging from the systematic examination of these research questions provide an important theoretical contribution, either specifically to the actor engagement role literature or more broadly to the general actor engagement literature. While actor engagement has been studied extensively in business contexts, its potential to drive change in collaborative efforts in domains such as healthcare, education and environmental sustainability remains underexplored, whereas engagement in these contexts requires a comprehensive understanding of how engagement can be facilitated (Alexander et al., 2025). The findings of this study contribute to closing these gaps by providing insights into how actor engagement practices can be understood and managed within collaborations.

Firstly, given that the literature review indicated the absence of a unified conceptual understanding of actor engagement roles, we developed a generic actor engagement role framework, thereby contributing to conceptual clarity by systematically delineating their characteristics. The framework, grounded in the actor engagement literature, defines actor engagement roles based on two primary components: engagement disposition and activity of engaging. Engagement disposition encompasses the degree of tendency, capacity and their valence. Activity of engaging captures the intensity of engagement in terms of the amount and duration of invested resources. Furthermore, the different dimensions are specified with respect to the kind, form and activeness of resources, and valence is also considered. In addition, the framework recognises connectedness as an emergent consequence of variations in engagement disposition and activity. Connectedness is thus treated as an outcome of configurations of engagement disposition and activity.

Importantly, the framework differs from existing teamwork and co-creation role typologies, which predominantly classify actors according to observable behaviours, task contributions or functional responsibilities. In contrast, our framework conceptualises roles as configurations of engagement disposition (capacity and tendency) and the activity of engaging. This distinction enables the identification of actors who may display similar behavioural patterns yet differ substantially in their underlying willingness or ability to invest resources. For example, low contribution may result either from limited capacity despite a positive tendency to engage or from substantial capacity combined with low willingness to invest resources. Although these actors may appear similarly disengaged from a behavioural perspective, they generate different engagement dynamics and require different managerial interventions. Thus, the framework explains why similar contribution patterns can produce different engagement dynamics, a distinction that would remain obscured in role classifications based solely on observable behaviour.

This detailed conceptualization allows a more nuanced understanding of how diverse patterns of resource investment are reflected in actor engagement roles. As the framework draws on the actor engagement concept, it can be used to develop typologies that refer to both individual and collective actors. Due to its generic nature, the framework can be applied across diverse contexts to identify actor engagement roles. Moreover, it offers a high degree of differentiation, which allows it to be used for the operationalization of actor engagement roles and thus provides a foundation for future measurement development.

Secondly, applying the theoretical framework as a sensitising device to collaborations in a student setting, we generated an actor engagement role typology and constellation by identifying eight actor engagement roles across three overarching categories. This typology is the first to capture not only actors' resource investments and engagement dispositions but also the emergence of connectedness as a consequence of their varying configurations, establishing a constellation for collaborative contexts with similar conditions. In this regard, the typology and constellation also serve as an important foundation for future advancement in this area, in particular the development of measurement instruments. The results derived from our typology and constellation advance the knowledge of actor engagement roles and actor engagement in three specific ways: Our typology focuses on individual actors, it accounts for negatively associated roles, and it is the first typology to demonstrate that roles may share similar engagement practices while differing in the intensity of their manifestation.

The typologies in the current literature primarily focus on (eco)system or network levels, overlooking that even in institutional cooperation for specific projects, individuals representing the institution interact with one another (Alexander et al., 2018). Consequently, understanding roles requires consideration at both institutional and individual levels. This is important since the engagement of individuals has a crucial impact on the outcome of the collaboration, highlighting the need to understand, in particular, the individual actor's engagement disposition to enhance their engagement. Accordingly, our typology offers a relevant contribution to the actor role literature by generating roles for actors in collaborations with specific attributes at the individual level.

Our generated insights contribute to a better understanding of the “dark side” of engagement (Li et al., 2017; Bowden et al., 2017). While actor (engagement) role typologies typically do not identify roles with negative engagement practices and impact, our work shows that actor engagement roles with negatively valenced tendencies and negatively valenced resource investment specifically affect the emotional engagement of other actors in a negative way, with potential negative impacts on their behavioural resource investment as well. Moreover, valence duality effects could also be observed across our roles. For example, the resource investment of impeding resource investors positively influences the behavioural engagement of collaborating actors while simultaneously negatively affecting their emotional engagement. Valence duality in engagement remains underexplored in the literature (Bowden et al., 2017); however, our differentiated conceptualization of engagement provides a foundation for better understanding such phenomena, and our typology offers empirical illustrations of these effects. Considering the implications of negative-valenced engagement on individual well-being and on collaborative outcomes (Nakintu et al., 2026), understanding such negative influences is crucial.

Our typology shows that the roles of activated resource investors exhibit the same engagement practices as the more extreme roles of leading resource investors and impeding resource investors, but in a mitigated form. This conceptualization provides a theoretical basis for understanding potential role transitions. While our empirical data capture actors' retrospective accounts within bounded collaborative projects rather than longitudinal observation across extended periods, the differentiated configurations identified here suggest how shifts in tendency or capacity may give rise to changes in role manifestation. For example, temporary high and supporting resource investors may, through targeted capacity enhancement (knowledge, experience, skills), be encouraged to assume leading roles. Conversely, deprived or reluctant resource investors may, due to frustration, lack of recognition or excessive workload, shift toward impeding roles. Accordingly, our typology provides a foundation for identifying subtle role differences and for assessing the potential or risk of role transitions by influencing these characteristics.

Thirdly, our findings offer an important contribution to the actor engagement literature by advancing the understanding of connectedness. Firstly, we provide a stronger theoretical foundation by defining connectedness as an outcome of variations in the characteristics of actor engagement roles, including engagement disposition and activity. Secondly, we empirically illustrate actor-induced engagement dynamics in collaborative settings through our typology and constellation. In doing so, we highlight the interdependence of actors within service (eco)systems. To date, connectedness has been considered in only one engagement role typology (Flaig and Ottosson, 2022), where it is treated as a characteristic of the role itself. In contrast, our approach defines roles based on varying manifestations of engagement disposition and activity of engaging, while conceptualizing connectedness as an outcome of these manifestations. Connectedness is thus inherently present, as actor engagement is fundamentally defined as occurring through interaction with connected actors (Brodie et al., 2019); however, connectedness is not a characteristic of the role itself. Therefore, our approach exhibits a higher degree of conceptual consistency with the definition and underlying conceptualization of actor engagement.

Lastly, the use of the actor engagement perspective, understood as a micro-foundation for value co-creation and value co-destruction, provides a complementary lens for understanding macro-level phenomena such as value co-creation or value co-destruction through a deeper comprehension of micro-level processes, in this case, resource investment (Storbacka et al., 2016; Lumivalo et al., 2024). The actor engagement perspective does not focus on outcomes but rather on the underlying processes, thereby enabling the consideration of phenomena that are not directly observable. This aligns with and contributes to research on micro-foundations, which, although not a new approach in the service and marketing literature, remains in its infancy in terms of conceptual and empirical development (Conduit and Chen, 2017; Lumivalo et al., 2024).

From a practical perspective, this paper offers two principal contributions: the first relates to the generic actor engagement role framework, while the second concerns the empirically derived typology and constellation of actor engagement roles.

Firstly, the actor engagement role framework can be particularly valuable for managers or facilitators, for instance in intra- or inter-organizational projects, as it enables a more nuanced understanding of individual project team members' engagement and supports facilitating and sustaining engagement over time across collaborating partners. In contrast to the typology, the framework is more neutral, as it does not prescribe predefined categories that assign positions within a social or collaborative structure. Therefore, it can be used to analyse project participants' engagement from an external perspective as well as a tool for self-evaluation. More specifically, the framework supports structured comparisons between self-assessed and externally assessed engagement practices, facilitating the identification of misalignments as well as under-recognized or underutilized engagement capacities. On this basis, it can inform targeted facilitation and capacity development measures, such as differentiated role design, tailored support interventions or the reallocation of tasks and responsibilities in accordance with actors' engagement disposition. For example, when a group member exhibits low willingness to engage, a leader may initially infer a generally low disposition. However, if self-assessment indicates that the actor possesses high capacity, this information can be used to identify mechanisms for altering the actor's tendency and activating available resources. For instance, our study shows that one interviewee shifted from a reluctant resource investor to a temporary high resource investor after learning that the final presentation would take place in a start-up centre with relevant stakeholders serving on the jury, thereby influencing the tendency to invest resources.

In future longitudinal applications, the framework could support tracking engagement trajectories and identifying early signs of decline or stagnation. We therefore recommend using it as an assessment tool at the beginning of collaborations and at regular intervals, depending on their duration. By considering both positively and negatively valenced engagement dispositions and activities, the framework also enables the early detection of emerging negative engagement and timely corrective action. This is particularly important given the potentially detrimental effects of negatively valenced engagement on collaborative processes (Nakintu et al., 2026).

For individual actors, the framework functions as a reflective self-regulation tool, supporting them in understanding and identifying actions to enhance their engagement. This contributes to collective goals and is associated with higher individual well-being (Virlée et al., 2020). This applies, for example, to supporting resource investors, who exhibit a high tendency and motivation to contribute but lack the necessary capacities, potentially leading to reduced self-efficacy and inadequacy. This can be addressed by organizing micro-workshops led by high resource investors or pairing supporting resource investors with temporary high resource investors to facilitate knowledge transfer. Over time, this enables them to build capacities, make meaningful contributions and enhance their well-being.

The second area in which our paper offers a practical contribution concerns collaborative processes, such as grassroots initiatives, emerging social movements, newly formed non-profit organizations and established NPOs, NGOs and similar organizations that rely particularly on volunteering, where our actor engagement role typology and constellation can be applied. These contexts share similar characteristics with our student setting and engagement is especially relevant in such environments. However, the transferability of these insights should be considered with caution, as contextual conditions vary across settings and may influence their applicability.

The typology may have potential as a peer-assessment instrument to support the alignment of roles and positions with suitable candidates, for example in leadership selection. For instance, in a newly established climate action group whose members already possess extensive expertise and experience, a leader primarily acting as an accelerating resource investor may be sufficient. In this case, the leader does not need to provide substantial content input but should facilitate collective organization and coordinated action by contributing resources such as project management, coordination and communication skills. However, if an association forms in the field of human rights with highly motivated members but limited expertise, a leader assuming the role of a high resource investor becomes essential. Such a leader must provide knowledge, experience, guidance and support while transferring expertise to other members. As our study shows that high resource investors can also diminish others' engagement, collaborations with such leaders should ensure that they act as catalysts for peer learning. This can mitigate potential negative effects and leverage their engagement to increase the capacity of supporting, deprived and scarce resource investors.

The typology and constellation may have potential not only for newly formed groups, where they could support the alignment of roles with suitable members, but also for established volunteer-based organizations. In such contexts, the framework could support role allocation and the management of volunteer engagement. For example, many NGOs and NPOs face challenges in recruiting successors for leadership positions, which are often held by older members who must eventually step down. Our typology could help identify temporary high resource investors who, through targeted support such as training and development, may be encouraged to assume high resource investor roles. Supporting resource investors could, for instance, gain experience by temporarily accompanying leadership positions in co-leadership arrangements, learning not only to organize tasks but also groups and activities, enabling them to later act as accelerating resource investors. Overall, our findings suggest that leadership can be effectively shared between individuals with complementary knowledge and skills. Co-leadership or dual leadership structures may therefore help reduce individual burdens, increase willingness to assume leadership responsibilities and address succession challenges in NGOs and similar organizations.

Furthermore, many NGOs and NPOs have a large pool of supporters but relatively few who actively contribute on their own initiative. Volunteer engagement departments could particularly benefit from our differentiated view of activated resource investors to design targeted measures that mobilize the latent potential of the broader supporter base.

Based on the typology, temporary high resource investors could be engaged in projects or activities through general emails calling for participation. For actors with the relevant capacities, this low-threshold activation is generally sufficient to prompt contribution. To leverage the potential of reluctant resource investors, however, actors should be approached personally, with concrete opportunities for recognition and reputation highlighted to influence their tendency and encourage resource investment. Personal outreach via telephone or in-person (online or offline) meetings can therefore be an appropriate means of activation.

Furthermore, our findings indicate that the high engagement of high resource investors can elevate the engagement of other actors. Events or joint working sessions should therefore be organized to leverage this effect and encourage the engagement of activated resource investors. This crossover effect occurs particularly when there is a high degree of interaction among actors (Costa et al., 2014). Accordingly, large-scale events that enable interaction among many individuals provide valuable opportunities to activate the engagement of activated resource investors at scale. This can foster collective engagement, which is highly relevant for achieving shared goals, as in volunteer-based NGOs such as Fridays for Future or Greenpeace. Activation processes may transform initially individual engagement into shared engagement among actors, thereby constituting collective engagement (Kleinaltenkamp et al., 2019).

To assess volunteers' engagement roles, NGOs and similar organizations could use our typology as a basis for a questionnaire during registration and/or at regular intervals, either digitally or at events, to capture volunteers' engagement disposition and activity and enable appropriate classification and adjustment of measures. Storing results in a digital database can streamline the work of engagement departments and enhance operational efficiency.

Several limitations should be acknowledged alongside the contributions presented. Firstly, participation was voluntary, potentially biasing the sample toward more committed students and limiting insights into less committed perspectives on actor engagement roles. Nonetheless, our research identifies roles that negatively influence collaborative processes, addressing a gap in existing typologies. Further research should examine actor engagement roles with negative manifestations, for example in controversial collaborations such as transition projects or emerging anti-climate or anti-immigration movements. A closer integration with the macro-level perspective of S-D logic could further explore actor engagement roles in value co-destruction processes and the interdependence of co-destruction and co-creation patterns (Lumivalo et al., 2024).

Secondly, both the framework and typology are based on a highly differentiated conceptualization of engagement practices. This nuanced understanding, together with the insight that activated resource investors exhibit practices similar to more extreme roles, can support the exploration of mechanisms that encourage transitions between roles. While we briefly addressed role transitions, future studies should examine this phenomenon in greater depth, for example through longitudinal or real-time observational designs. Such designs would enable the investigation of how engagement roles emerge, stabilize or shift over time.

Thirdly, this study aimed to conceptualize an engagement role typology and constellation applicable to collaborative contexts characterized by the absence of pre-existing hierarchies, roles and monetary exchange. While some characteristics are context-specific, such as the lack of language skills as a frequent resource gap in the examined international student groups, this setting was selected to facilitate the examination of role formation with a focus on actors' engagement, while acknowledging that other factors also influence role formation and enactment. Validation across contexts where engagement is central to collaboration, such as emerging social movements, newly formed NGOs, NPOs or associations, is encouraged. Such research could also examine additional factors influencing role formation beyond engagement and quantitatively validate the typology.

The empirical typology and constellation developed here derive from temporary collaborations without formal hierarchy and limited prior relational history. In more structured organizational settings, authority, established role expectations and incentive systems may shape engagement configurations differently. Future research should examine how the framework manifests under varying institutional and structural conditions. Moreover, as not all roles emerge in every collaboration, further studies should investigate which role constellations occur across contexts and which factors influence their emergence.

Beyond contextual transferability, a related conceptual limitation concerns the extent to which identified roles reflect engagement configurations alone rather than the influence of structural factors such as gender, age, cultural background and prior relational history. Although the student context was selected to minimize pre-existing hierarchies and roles, these factors cannot be fully excluded as contributors to role formation and enactment. For instance, cultural norms around assertiveness or deference may shape an actor's tendency to invest resources in ways that are difficult to distinguish from engagement disposition itself. Future research should therefore examine the relative influence of structural and engagement-based factors on role formation through designs that allow these influences to be compared or isolated.

Lastly, data analysis revealed the crucial role of team, project or collaboration leaders. Even in processes without predefined hierarchies, leaders significantly influenced overall success. Future research should examine how resource investment leaders contribute to the emergence of collective engagement, as highly engaged actors are pivotal in fostering it (Van Mierlo and Bakker, 2018). Moreover, interviews with individuals holding a leading resource investor role indicated that these actors often assumed leadership functions in other collaborations as well. Future research could therefore investigate whether high engagement is associated with specific character traits or dispositional factors, and which contextual conditions influence when actors assume leadership positions across collaborations. This could build on existing research examining the nature and processual character of human actors in service literature (Koskela-Huotari and Siltaloppi, 2020).

The supplementary material for this article can be found online.