Abstract
Teams are an important resource that organizations rely on to create (i.e., ideation) and implement (i.e., implementation) innovative ideas. To advance distinct theorizing on ideation, implementation and their transit as crucial but different phases of innovation in teams, we collected qualitative data from 60 individuals working on innovation in teams and analyzed 235 situations critical for ideation and implementation. Our qualitative content analysis resulted in 1460 codes classified into 27 categories, which we analyzed with regard to their relative importance for ideation and implementation in teams. We integrated our results via an abductive approach using existing theory into a process model showing that ideation is driven by information processing and team climate, implementation is driven by action regulation and resources, and the transit phase is driven by a mindset change from deliberation to implementation. The interface between ideation and implementation reveals paradoxes that deserve explicit management.
Keywords
To promote innovations, organizations increasingly rely on teams (Ilgen et al., 2005; Mathieu et al., 2017), because teams “are a key source of creativity and innovation” (Rouse, 2020; van Knippenberg & Hoever, 2021, p. 49). Innovation, the “intentional introduction of new and useful applications, products, or procedures (Rietzschel et al., 2021, p. 129; West & Farr, 1990), consists of idea generation and idea implementation which involve two different, competing activities, exploration and exploitation (Rosing et al., 2011). This duality, known as ambidexterity (Rosing et al., 2011), is a key challenge for teams working on innovations (Anderson et al., 2014; van Knippenberg, 2017). Frequently innovation fails because novel ideas cannot be (successfully) implemented (Singh et al., 2021; West, 2003) 1 .
Despite the practical relevance of idea generation and implementation in teams for successful innovation, team innovation research often focuses on the ideation phase of innovation (Hülsheger et al., 2009; for an exception, see Byron et al., 2023). Particularly, idea implementation in teams, which could complement a full understanding of team innovation, has received little attention to date (van Knippenberg, 2017), making idea implementation and the “contingencies of the relationship between idea development and idea implementation . . . the most important understudied issue in team innovation research” (van Knippenberg, 2017, p. 211). Methodologically, this focus might be due to the insufficient psychometric distinctiveness of both constructs (van Knippenberg, 2017), which could be solved by diving more deeply into idea generation and idea implementation using qualitative research (Rouse & Pratt, 2021). Theoretically, this focus might be due to a lack of motivational team-level theorizing at the interface between idea generation and idea implementation, which could be overcome by integrating different strands of respective individual level and team level theories.
In this study, we empirically disentangle the processes of idea generation and idea implementation in teams, shed light on their distinct underlying mechanisms and conditions according to an input-mediator-output logic (Kozlowski & Ilgen, 2006), and elaborate on the motivational and behavioral nature of the transit phase between idea generation and idea implementation. We interpret our results abductively, integrating various strands of cognitive, behavioral and motivational theories at the individual and team level: We present ideation in teams as an information processing path according to van Knippenberg’s (2017) model of team innovation. We further propose implementation in teams as an action regulation process according to action regulation theory (Frese & Zapf, 1994; Hacker, 1994; Zacher & Frese, 2018) and make sense of the transit phase between ideation and implementation using the Rubicon model of action phases (Heckhausen & Gollwitzer, 1987; also see Keller et al., 2020).
Combining these results to a new model of team innovation, we respond to the call for theory-building (Anderson et al., 2014) focusing on how the innovation process unfolds (Rouse & Pratt, 2021). We contribute to team innovation research by peeling out idea implementation as a distinct, action regulation path besides the information processing-based idea generation. By providing insights into the transit between idea generation and idea implementation with the application of individual-level, motivational and action regulation theories at the team level (Zacher & Frese, 2018), we offer impulses for an advancement of innovation research in specific (Rietzschel & Ritter, 2018) and ambidexterity theory in general (Zhou et al., 2023).
By conducting our study in a field context, we address the “important omission in research in team creativity and innovation that research seems to hardly concern such contexts in which the distinction between idea generation and idea implementation is worth studying” (van Knippenberg & Hoever, 2021, p. 60). The field setting of our research also enhances the practical validity and relevance of our results, which offer concrete indications regarding the specific processes and factors needed to support different phases of the innovation process.
In the following, we do not provide an exhaustive review but rather summarize representative literature that is most relevant to the empirical model we seek to develop: We first provide an overview of theory and research on predictors of team innovation and show that they often focus either on specific predictors of ideation or predictors of implementation. We then describe previous research on predictors of ideation and implementation and address the problem that this research often refers to individual-level innovation and does not consider the fact that innovation often takes place in teams. We conclude that, although predictors of individual-level ideation and implementation have been identified, research on team innovation often neglects this distinction.
Predictors of Team Innovation
Team innovation can be defined as the “extent to which a team, as a whole, develops useful new ideas and converts them into outputs” (Zhang et al., 2022, p. 1654; also see Drach-Zahavy & Somech, 2001). When it comes to team outcomes, such as team innovation, team interaction processes are of utmost importance as antecedent variables (cf., Drach-Zahavy & Somech, 2001; Salas et al., 2015). Team processes capture “how team members combine their individual resources, coordinating knowledge, skill, and effort to resolve task demands” (Kozlowski & Ilgen, 2006, p. 77). Team process variables, such as communication, support for innovation, task orientation, collaboration, team reflexivity, team information elaboration, team intrinsic motivation and team potency, were meta-analytically found to be stronger predictors of team innovation than team structural variables (e.g., team size), task structure variables (e.g., task interdependence) or organizational support variables (team leadership) (Byron et al., 2023; Hülsheger et al., 2009). Besides team processes, recent models of team innovation also consider team climate variables as moderating influence when it comes to team innovation (van Knippenberg, 2017).
Predictors of Ideation and Implementation
Innovation consists of idea generation and idea implementation. That is, generating a good idea may only result in a successful innovative product if it is also implemented accordingly 2 (Bledow et al., 2009; West, 2003). There is much debate in management research about the sequential ordering of ideation and implementation: Most research has shown that the innovation process is a non-linear, cyclical process (Anderson et al., 2004, 2014; Rosing et al., 2011; West, 2002) that alternates between idea generation and idea implementation (Anderson et al., 2014; Rosing et al., 2011). Idea generation and idea implementation are positively related to each other (for a meta-analysis, see Sarooghi et al., 2015), but involve different activities and are guided by different leadership styles: Idea generation requires exploration (variation, experimentation, and flexibility) and is best guided by authentic, empowering, and entrepreneurial leadership. Idea implementation implies exploitation (efficiency and execution), facilitated by transactional (contingent reward) and supportive leadership (Bledow et al., 2009; Lee et al., 2020; March, 1991; Rosing et al., 2011).
Regarding individual innovation, there is evidence that ideation and implementation might be driven by different antecedents (Hammond et al., 2011). Contextual factors such as autonomy, supervisor support, and innovation climate correlate more strongly with implementation, whereas factors such as openness, intrinsic motivation, and extrinsic motivation have stronger relationships with ideation (Hammond et al., 2011). Singh et al. (2021) meta-analytically identified further predictors of innovation implementation in organizations, including transformational leadership, communication and collaboration, internal and external entrainment of functions and systems, leader and employee competency, absorptive capacity, availability of resources, and implementation climate.
Ideation and Implementation in Teams
Although predictors of individual-level ideation and implementation have been identified meta-analytically (Hammond et al., 2011), research on team innovation often neglects this distinction (van Knippenberg, 2017; van Knippenberg & Hoever, 2021). As reported by van Knippenberg and Hoever (2021), only a few studies differentiate between ideation and implementation regarding team innovation: Huang et al. (2017, p. 693) for example found that “idea generating benefits from higher diversity on team traditionalism, whereas idea implementing benefits from higher average levels of team traditionalism”. Somech and Drach-Zahavy (2013, p. 684) showed that “aggregated individual creative personality, as well as functional heterogeneity, promotes team creativity, which in turn interacts with climate for innovation” in predicting innovation implementation. Only when climate for innovation was high did team creativity foster innovation implementation (Somech & Drach-Zahavy, 2013). Byron et al. (2023) compared predictors of creativity versus innovation (i.e., idea generation vs. idea implementation) and found that team tenure had a stronger positive relationship with team innovation than with team creativity. West (2002) theorized on creativity versus innovation implementation in work groups and suggested opposite effects of external demands on creativity (inhibiting effect) and innovation implementation (facilitating effect).
These studies show patterns similar to results on individual ideation and implementation, that is, a stronger influence of contextual factors, such as climate for innovation or external demands on idea implementation than on idea generation. Regarding idea generation, these studies mainly focus on structural variables, such as team member characteristics (diversity, creative personality), neglecting team processes. Thus, these studies mainly illuminate contextual features and characteristics of innovative teams and organizations and neglect the behavioral repertoire of teams. A more general theoretical framework that could serve as a basis for future systematic team innovation research is thus needed.
Taken together, teamwork is essential for innovation (Wuchty et al., 2007). Innovation is an ambidextrous phenomenon consisting of ideation and implementation of ideas (Hammond et al., 2011; Potočnik et al., 2022; West, 2003). However, in research on team innovation there is a lack of empirical and theoretical differentiation between idea generation and idea implementation (e.g., Hülsheger et al., 2009; van Knippenberg, 2017), and their interplay (e.g., Singh et al.’s (2021) focus on implementation; Liu et al.’s (2016) focus on creativity), which would be important for the further systematic and aligned development of the research field. We therefore address the research questions: How does the team innovation process unfold? Are idea generation and idea implementation in teams driven by different processes?
We addressed these questions with a qualitative research methodology and an abductive approach to data interpretation. The qualitative methodology allowed us to disentangle the specifics of idea generation and idea implementation and to overcome the limitations of quantitative research methods that end up mixing ideation and implementation due to high intercorrelations (Rouse & Pratt, 2021). Following the logic of abduction (Sætre & Van de Ven, 2021), we generated theoretical explanations for the identified ideation process in teams using the knowledge integration/team climate model of team innovation (van Knippenberg, 2017), for the implementation processes using action regulation theory (Frese & Zapf, 1994; Hacker, 1994; Zacher & Frese, 2018) and for the transit phase between ideation and implementation using the Rubicon model of action phases (Heckhausen & Gollwitzer, 1987; also see Keller et al., 2020). In this way, we brought together different strands of cognitive, motivational and behavioral individual level and team level theories and formulated a new model of team innovation, which sheds light on the distinction and transit between ideation and implementation and thereby also offers impulses for innovation research in specific and ambidexterity theory in general.
Method
As research on distinct predictors of team ideation and team implementation is scarce and quantitative methods are only of limited use due to the difficulty of distinguishing ideation and implementation psychometrically (see van Knippenberg, 2017), we chose an exploratory qualitative approach (Miles et al., 2014) to data collection and an abductive approach (Sætre & Van de Ven, 2021) to data interpretation. Moreover, qualitative methods are well-suited to “incorporate the specificity and realism that are critical for understanding creativity dynamics within the contexts of organizations” (Rouse & Pratt, 2021, p. 310). We first identified antecedents of idea generation and idea implementation in teams in our data. We then generated plausible explanations for our findings using the knowledge integration/team climate model of team innovation (van Knippenberg, 2017), action regulation theory (Frese & Zapf, 1994; Hacker, 1994; Zacher & Frese, 2018) and the Rubicon model of action phases (Heckhausen & Gollwitzer, 1987; also see Keller et al., 2020) (which will be described in more detail below), and integrated the identified categories into a comprehensive process model.
Sample and Participant Selection
We collected data from 60 participants (51.7% female, mean age = 32.4 years, mean work experience = 5.0 years, 70.0% with leadership or project responsibility, 73.3% with university degree) from different industries (e.g., 20.0% health, 16.7% engineering, 8.3% real estate, 8.3% marketing, 6.7% research, 6.7% management, 6.7% entertainment, 6.7% education, 5.0% finance, 5.0% technology). We conducted semi-structured face-to-face or phone interviews with 21 participants. The remaining 39 participants answered the interview questions via an online questionnaire. Data saturation (i.e., further interview data did not generate any new aspects that went beyond those identified in the previous interview data, O’Reilly & Parker, 2013) was reached after ten face-to-face/phone interviews, which is a criterion for determining sample size (O’Reilly & Parker, 2013). However, we conducted additional face-to-face/phone interviews to underpin our category system. We then collected further data via online questionnaires to obtain data sufficiency (Shaheen & Pradhan, 2019), that is, a larger base of codes that allowed us to determine the relative frequencies of our categories. Following the logic of purposeful sampling (Palinkas et al., 2015; Patton, 2002; Suri, 2011), we defined the inclusion criterion for participant selection in advance, that participants (had) worked on innovation (including ideation and implementation) in teams.
Face-to-face/phone interview candidates who were considered information-rich cases regarding team innovation (i.e., individuals who matched our inclusion criterion) and could thus contribute to our specific inquiry (purposeful sampling, see Patton, 2002) were identified in different ways: Eight persons were identified by a contact person from an existing cooperation with an engineering company within a large research project. Five persons were identified through personal relationships between one of the authors and a technology company, and eight persons via relationships between a research assistant and further companies linked to innovative tasks and through an Internet research regarding innovation teams. As everyone involved in the participant recruitment process was aware of the inclusion criterion, all participants invited to participate in the study met it. We achieved maximum variation by selecting individuals from different industries and with different demographic characteristics (Miles et al., 2014). In order to test our procedure as well as the interview guide, we conducted two pre-test interviews, one of which was included in the data analysis as the interview candidate fit our inclusion criterion.
For the online questionnaire, we identified participants based on their jobs and tasks and contacted them via email. For this purpose, available knowledge about people’s areas of responsibility stemming from personal contacts was utilized. In this way, people who potentially met the inclusion criterion of the study could be targeted. In the online questionnaire, we again communicated that a prerequisite for participation was that participants worked in teams on tasks that required ideation and implementation. Moreover, we asked participants to suggest further potential participants (snowball sampling, Shaheen & Pradhan, 2019). Here, too, we strove for maximum variation in order to reach data sufficiency (Shaheen & Pradhan, 2019). Due to snowball sampling, response rates could not be determined.
Data Collection
Interview Procedure
Participants for the face-to-face/phone interviews were first contacted via email to provide them with background information about the research as well as the interview procedure. Participants provided informed consent prior to the interviews. The interviews lasted 39.9 min on average (Min = 22 min, Max = 56 min). The interviews were digitally recorded. Interviews were conducted by one author of this paper and one research assistant. Participants for the online questionnaire were contacted via email and received the link to the questionnaire. On the first page of the questionnaire, participants provided informed consent. The study was carried out in accordance with the recommendations of the Ethical Principles of Psychologists and Code of Conduct by the American Psychological Association (2017) and with the Declaration of Helsinki (World Medical Association, 2013). Participants did not receive compensation or incentives for their participation. Data collection was conducted in German language.
Interview Guide and Online Questionnaire
For the face-to-face/phone interviews, we used a semi-structured interview guide (Flick, 2007) (see Appendix A). We started the interview by explaining the topic and purpose of the interview. Then, a short explanation of ideation and implementation was provided to develop a common understanding of innovation. We then asked for critical incidents from participants’ work experience in innovation situations and respective inhibiting and facilitating factors within these situations (Flanagan, 1954). We also asked for general, typical team characteristics and teamwork aspects that were important for ideation and implementation; however, our focus for the analyses reported in this paper was on the critical incidents.
Similarly, the online questionnaire started with a short explanation of ideation and implementation, and asked for critical innovation situations and respective inhibiting and facilitating factors in these situations. Specifically, participants were asked to describe (1) a successful ideation situation in their team, (2) an unsuccessful ideation situation in their team, (3) a successful idea implementation situation in their team, and (4) an unsuccessful idea implementation situation in their team. For each critical incident, we asked participants to describe the situation as well as the team behaviors and conditions that promoted (successful incidents) or impeded (unsuccessful incidents) success within the situation. At the end of the questionnaire, we collected demographic data.
Data Set
In total, we analyzed 235 situations (59 successful ideation situations, 57 unsuccessful ideation situations, 60 successful implementation situations, 59 unsuccessful implementation situations), resulting in a total of 1460 codes. The initial coding was conducted by one author of this paper in close dialogue with a second author and a research assistant. A random subsample of 42% of these codes, which covered successful and unsuccessful ideation and implementation situations equally, was double-coded by a third author of this paper. Initial interrater agreement before discussion was 93%. Disagreements were discussed and a consensus of 97% was reached after discussion. After this process, the initial coder reviewed and, if necessary, adjusted the remaining codes according to the consensus reached in the discussion and to ensure that all statements were double-checked in this way.
Data Analysis
Audio recordings from the face-to-face/phone interviews were transcribed verbatim in accordance with pre-defined transcription rules (Kuckartz et al., 2007). Then, following qualitative content analysis principles by Mayring (2015), we analyzed data from both the interviews and the online questionnaire in five steps.
Category System: Factors Influencing Ideation and Implementation in Teams.
Note. Supplements in parentheses and italics reflect the expression of the category in unsuccessful ideation and implementation situations. The table logic is illustrated by the following example: The statement “Simply too many different characters came together, so that the whole team could not really work together well.” (#14) (see column “example”) refers to “too much team diversity regarding personalities” (see column “category content and description”) and was thus categorized as “too much team diversity” (see column “category”). Team diversity together with team size represent the higher-level theme “team composition” (see column “higher-level theme”).
Results
Our final category system consisted of 27 categories (see Table 1, second column), which were assigned to 9 higher-level themes (see Table 1, first column). For definitions of each category as well as prototypical examples, see Table 1 (third and fourth columns). The category system fit successful and unsuccessful ideation and implementation situations in teams. Absolute frequencies of categories are shown in Figure 1. Regarding ideation, the categories participative safety, motivation and volition, information absorption, goal orientation, and open-mindedness were mentioned most frequently. Regarding implementation, the categories resources, coordination, motivation and volition, management support and monitoring were mentioned most frequently. Absolute frequencies of each category regarding ideation and implementation. Note. Multiple mentions per person and per category were allowed.
Figure 2 illustrates differences in the relative frequencies of categories for ideation versus implementation. For each category, we subtracted the relative frequency of implementation-related mentions from the relative frequency of ideation-related mentions. Differences were statistically tested for significance using χ2 tests (for details, see Appendix B). Positive difference scores indicate that the category was mentioned more often for ideation than for implementation. Negative difference scores indicate that the category was mentioned more often for implementation than for ideation. For example, monitoring was mentioned 20 times for ideation, which corresponds to 20/768 = 2.60% of all ideation mentions. Monitoring was mentioned 42 times for implementation, corresponding to 42/692 = 6.07% of all implementation mentions. The difference between 2.60% and 6.07% (−3.47%, which is statistically significant) shows that monitoring was mentioned more often for implementation than for ideation. Non-significant difference scores indicate that the category was mentioned equally often for both ideation and implementation. We assumed that categories mentioned more frequently were cognitively more accessible for interviewees and thus more relevant with respect to the specific situations described. Therefore, we considered the relative frequency of mentions as an indicator of the relevance of each category for ideation and implementation in teams. By considering frequencies when summarizing our data we were able to focus on our key findings and identify emergent patterns with greater clarity (Dey, 1993, according to Sandelowski, 2001). Differences in relative frequencies of categories for ideation versus implementation. Note. The figure shows the differences in the percentage of statements referring to each category between ideation and implementation. Positive difference scores indicate that the category was mentioned more often for ideation than for implementation. Negative difference scores indicate that the category was mentioned more often for implementation than for ideation. Non-significant differences scores indicate that the category was mentioned equally often for both ideation and implementation. N mentions ideation = 768. N mentions implementation = 692. * Difference is statistically significant, p < .05. (*) Difference is marginally significant, p < .10. n.s. Difference is not statistically significant, p > .10. n.a. Test statistics are not interpreted because expected frequencies are too low (N </ = 5).
Model Formulation and Abduction
To integrate our categories into a model, we first selected categories that were mentioned significantly more often regarding ideation compared to implementation (upper part of Figure 2). Then we selected categories that were mentioned significantly more often regarding implementation compared to ideation (lower part of Figure 2). Finally, we selected the 50% of categories that were mentioned most frequently overall (upper part of Figure 1). Having selected the respective categories, we arranged them according to existing input-mediator-output logics (Mathieu et al., 2008): Categories referring to team composition and team member characteristics were placed as inputs (cf. van Knippenberg & Hoever, 2021). Emergent states and climate categories as well as support-related categories were placed as moderators (see van Knippenberg, 2017), which can also have direct effects on process categories (see Coultas et al., 2014; van Knippenberg & Hoever, 2021). Information processing and action regulation categories were placed as mediators (see, e.g., Mathieu et al., 2008).
Clustering and displaying the categories according to this logic revealed two different paths, one path leading to ideation and one path leading to implementation, as well as a transit phase between ideation and implementation. Following the abduction method (Sætre & Van de Ven, 2021), we then generated plausible explanations for the different paths and phases and identified them in information processing theory (cf. van Knippenberg, 2017), which we applied to explain the ideation path, in action regulation theory (Hacker, 1994; Zacher & Frese, 2018), which we applied to explain the implementation path, and in the Rubicon model of action phases (Heckhausen & Gollwitzer, 1987; also see Keller et al., 2020) to make sense of the transit phase between ideation and implementation.
Information Processing Theory
The knowledge integration/team climate model of team innovation (van Knippenberg, 2017) combines a knowledge integration perspective and a team climate perspective on team innovation. In an information processing logic, team diversity can serve as an informational resource. Teams’ “different ways to capture [such] informational resources predict team information elaboration” (p. 225), which then leads to team innovation. How informational resources are used, and information is elaborated depends on climate influences and motivators of information elaboration, such as team climate for innovation (Anderson & West, 1998), cooperative goals and shared identity (van Knippenberg, 2017).
Action Regulation Theory
Action regulation theory (Frese & Zapf, 1994; Hacker, 1994; Zacher & Frese, 2018) describes actions as a sequence of different phases: In the goal development phase, goals are assigned externally or generated internally through self-selection. In the orientation phase, individuals search for action-relevant information to detect opportunities and constraints on successful goal attainment. In the phase of plan generation and plan selection, individuals develop different plans and select those most likely to lead to goal attainment. In the phase of execution and monitoring of the plan, individuals compare their goal, plan and actual behavior. In the phase of feedback processing, individuals process “information from the environment about their current performance or progress toward attaining the goal” (Zacher & Frese, 2018, p. 125). According to action regulation theory, individuals are not bound to a strict, linear progress of these phases; instead, they can move back and forth between phases or even skip phases – especially if later phases prompt goal changes. Additionally, in some occasions, like when goals evolve, some phases may need to be repeated, while in others, such as when having multiple goals, some phases might occur simultaneously (Zacher & Frese, 2018).
The Rubicon Model and Mindset Theory of Action Phases
According to the Rubicon model of action phases (Heckhausen & Gollwitzer, 1987; also see Keller et al., 2020), a course of action includes four phases: At the beginning, there is a deliberation phase where individuals assess their wishes to make a binding decision on which wish to pursue as a goal (pre-decisional phase). Making this decision means crossing the Rubikon. Having crossed the Rubikon, individuals enter a planning phase where concrete strategies for achieving the goal are developed (pre-actional phase). In the subsequent phase, these plans are implemented (actional phase), and finally, achieved outcomes are evaluated (post-actional phase). The Rubicon model has been further developed into the mindset theory of action phases (e.g., Gollwitzer, 2012), which conceptualizes crossing the Rubikon as a transit from a deliberative predecisional mind-set with realistic, objective, and broad information search and information processing to a postdecisional implementation mind-set focused narrowly on realizing the goal (Heckhausen, 2020).
In sum, our model was formed as follows: The contents of the model stem from our empirically identified categories and their empirically identified affiliation to idea generation, idea implementation, or the transit phase between ideation and implementation. The process-structure of the model was derived from existing input-mediator-output logics (Mathieu et al., 2008). The theoretical meaning of the processes was abductively interpreted according to existing theory on team innovation (van Knippenberg, 2017), action regulation (Frese & Zapf, 1994; Hacker, 1994; Zacher & Frese, 2018), and action mindsets (e.g., Gollwitzer, 2012).
Discussion
An Information Processing–Action Regulation (ImpAct) Model of Team Innovation
Based on our categories and the abductive methodology we applied to them, we suggest an information processing–action regulation model of team innovation consisting of one path leading to ideation, one path leading to implementation, and an interface between the two processes representing a transit phase between ideation and implementation.
The ideation path is suggested to follow an information processing logic (cf. van Knippenberg, 2017): A diverse team composition (regarding expertise, experience, personality, and perspectives) serves as an informational resource. Team members then process information, that is, they differentiate, elaborate and integrate information. The quality of information processing depends on the team climate categories open-mindedness and participative safety, which can also have a direct impact on information processing. Ideation in teams is thus suggested to depend on the quality of information processing (Figure 3, upper part). An information processing–action regulation (ImpAct) model of team innovation. Note. Curved arrows symbolize iterative shifts between idea generation / ideation and idea implementation.
The implementation path is suggested to follow an action regulation logic (cf. Frese & Zapf, 1994; Hacker, 1994): Team characteristics such as team members’ expertise serve as an informational resource. Team members then transform information into tangible outcomes through action regulation processes of prototyping and monitoring. The quality of action regulation depends on support for innovation, that is, management support, instrumental resources and team members’ entrepreneurial attitude, which can also have a direct impact on action regulation processes. A further input in the implementation path is the quality of the product/idea, which means that the course of the implementation phase also depends on the (preliminary) outcomes of the ideation phase. Implementation in teams thus is suggested to depend on the quality of (collective) action regulation (Figure 3, lower part).
We also found that a specific set of categories was relevant for both idea generation and idea implementation: Idea generation, although primarily driven by information processing and climate, also required specific aspects of support (autonomy, motivation and volition) and action regulation (strategy and planning, coordination), while implementation, although primarily driven by action regulation and support, also required specific aspects of climate (goal orientation) and information processing (information absorption, information sharing). We interpreted this set of variables as constituting the transit phase between ideation and implementation (see Figure 3, middle part). This transit phase between ideation and implementation is suggested to follow a logic of mindset change, caused by crossing the “Rubicon” (cf. Gollwitzer, 2012; although the Rubicon model and mindset theory of action phases are originally theories of individual goal pursuit, we apply their basic tenets to the team context abductively). The transit phase is characterized by both elements of predecisional deliberation (e.g., information absorption, information sharing), and elements of postdecisional planning and action (strategy and planning, coordination). The climate in this transit phase is characterized by goal orientation, motivation and volition, as well as by autonomy. Thus, transiting from ideation to implementation in teams is suggested to depend on the quality of action-related mindset change within the team (Figure 3, middle part).
Summary of Findings
Even though most innovations are worked on in teams (cf. Wuchty et al., 2007) and most innovations fail during the idea implementation phase (Baer, 2012; Klein & Sorra, 1996), research and theoretical reasoning on ideation versus implementation in teams is scarce (van Knippenberg, 2017; van Knippenberg & Hoever, 2021). To disentangle idea generation and idea implementation in teams and put research on team innovation on a sound theoretical basis, while drawing upon existing theoretical approaches, we qualitatively explored processes of idea generation and idea implementation in teams and their transit phase, theoretically explained them, and displayed them in a process model.
Answering our research questions, this ImpAct model of team innovation differentiates two paths of team innovation: an information processing path leading to idea generation and an action regulation path leading to idea implementation. Both processes unfold through an input-mediator-output logic with moderating influences of team climate (especially for ideation) and resources (especially for implementation). We also identified an overlap between the information processing path and the action regulation path, which indicates a transit phase between idea generation and idea implementation, requiring a mindset change from deliberation to implementation.
On the one hand, our results provide evidence for the distinctiveness of ideation and implementation because different theoretical paths for ideation versus implementation were revealed (contradicting van Knippenberg & Hoever, 2021). On the other hand, our findings support the notion that “creativity and implementation do not neatly proceed in a linear fashion (…) and, therefore, cannot be easily split into separate phases or stages” (Rosing et al., 2011, p. 957). Our data disclosed the cyclical nature of idea generation processes and idea implementation processes (cf. category “iterative processes”) and a substantial overlap between both processes in the transit phase.
Theoretical Contributions
Idea Generation and Idea Implementation – Two Distinct but Interconnected Paths Leading to Innovation
Although van Knippenberg and Hoever’s (2021, p. 59) review “does not reveal a pattern of different theory and findings for creativity as compared with innovation,” we identified different processes for idea generation and idea implementation in teams, that is, an information processing process and an action regulation process. Just as research argues, that idea generation and idea implementation involve different activities and principles (exploratory activities and principles such as variation, experimentation, and flexibility vs. exploitative activities and principles such as efficiency and execution, see Bledow et al., 2009; March, 1991; Rosing et al., 2011; Wilden et al., 2018), we identified specific ideation-related categories and specific implementation-related categories in teams. Whereas ideation was driven by information processing and climate aspects, implementation was driven by action regulation processes and resources.
Our results also show that idea generation and idea implementation are iterative processes, and teams often switch between the phases on the way to team innovation, which also underlines the close intertwining of the two phases (cf. Anderson et al., 2014; Bledow et al., 2009; Rosing et al., 2011; see also Georganta et al., 2020a, 2020b, for similar iterative team processes in the context of team adaptation). Due to this cyclical nature of ideation and implementation (analogous to the cyclical nature of exploration and exploitation, see Anderson et al., 2004; Anderson et al., 2014; Rosing et al., 2011), we displayed idea generation and idea implementation as two distinct processes with respective interfaces in our ImpAct model of team innovation. These interfaces (categories relevant for both ideation and implementation) might also be a reason for the previous difficulty separating the two paths. Through our qualitative work, we were able to precisely identify and depict the distinct aspects of each process and their overlap.
Getting from Ideation to Implementation – Getting from Information Processing to Action Regulation
With these interface categories relevant for both ideation and implementation (goal orientation, motivation and volition, autonomy, information absorption, information sharing, strategy and planning, and coordination), we can also address the problem of getting from idea generation to idea implementation, which we propose to be a transit phase between information processing and action regulation, requiring a respective mindset change from deliberation to implementation. Since ideation and implementation overlap in the aspects mentioned above, we interpret these aspects according to the Rubikon model as critical for the transit between ideation and implementation.
Paradoxes and Tensions in the Innovation Process
The interface categories also illustrate the ambidextrous nature of innovation: On the one hand, team members demand autonomy, that is, task-related leeway rather than restrictive parameters; on the other hand, they demand goal orientation, that is, a clear mission with clear goals to which team members are aligned. This structure-freedom paradox was also identified by Potočnik et al. (2022).
Taking a closer look at goal orientation, which also encompasses individuals’ identification with an idea, and motivation and volition, which encompasses persistence, we also identified a tension between a strong identification with one idea and persistence in pursuing it versus openness to new or different ideas, as indicated in the open-mindedness category. This tension shows parallels to the flexibility-persistence paradox, which claims that people need both cognitive flexibility and persistence in both the idea generation and idea implementation phases (Ivcevic & Nusbaum, 2017; Nijstad et al., 2010; Potočnik et al., 2022).
Goal orientation, especially alignment with common goals, is the starting point for a further tension we detected. On the one hand, teams need a strong alignment towards their common goals and need to be synchronized in that direction. On the other hand, they should use their diverse attributes and think divergently. Potočnik et al. (2022) refer to this tension as the paradox of cohesion and uniqueness.
We also found indications for the differentiation-integration paradox (cf. Potočnik et al., 2022). On the one hand, team members mentioned a need for information differentiation through brainstorming and divergent thinking. On the other hand, they mentioned a need to elaborate on and integrate information by summarizing, discussing and documenting ideas.
This “both/and” between structure and freedom, flexibility and persistence, uniqueness and cohesion, and differentiation and integration, is symptomatic of the ambidextrous nature of innovation. In our study, these paradoxes were especially located in the idea generation phase and in the transit phase between ideation and implementation.
Information Processing, Team Climate, Mindset Change, Action Regulation and Resources – Different Perspectives and Their Interplay
With our ImpAct model of team innovation, we build on van Knippenberg’s (2017) information processing/team climate perspective on team innovation. However, we identified an additional action regulation process leading to innovation implementation, which we clearly position as a distinct path to the information processing path with overlaps in the transit phase. We showed how team climate categories influence ideation and support-related categories influence implementation. Thus, going beyond van Knippenberg (2017), our model integrates information processing perspectives, team climate perspectives, mindset change, action regulation perspectives, and a resource perspective (support for innovation) and shows how information processing, climate, action regulation and resources work together as antecedents of team innovation.
Practical Implications
Our results provide practical implications for the management of ideation and implementation in teams and hint at how to switch between the phases of the innovation process: To support the ideation phase, teams could be trained to focus on information processing, to identify and access different sources of information (which can be facilitated by a diverse team composition), and to differentiate information by thinking divergently, such as in brainstorming sessions. Discussing and summarizing ideas within the team contributes to the elaboration and integration of promising ideas. These ideas then need to be documented. A positive team climate in which members are open to new perspectives, ideas, and change, trust each other, and feel psychologically safe to contribute their ideas can be helpful to support information processing.
To support the implementation phase, teams can be trained to focus on action regulation. Based on team members’ expertise and the quality of the developed ideas, the team can be trained to start prototyping ideas and to monitor its efforts by consolidating with stakeholders, seeking out feedback, checking the feasibility of prototypes, reflecting and learning. In the implementation phase, it is particularly important for teams to have members with an entrepreneurial attitude and have access to relevant resources and support.
Besides team training, leaders have the power to provide support and access to resources for implementing ideas (cf. Škerlavaj et al., 2014) and contribute to the development of an innovation-promoting team climate. They can show opening behaviors (giving room for own ideas, encouraging experimentation with different ideas) in the ideation phase (cf. information differentiation) and closing behaviors (monitoring and controlling goal attainment, see Rosing et al., 2011) in the implementation phase (cf. prototyping) to support the team innovation process. Although not directly addressed in our data, many of our categories are related directly (management support for innovation) or indirectly (goal orientation, participative safety) to leadership behaviors and functions.
Regarding the transit phase between ideation and implementation, a mindset change within the team has to take place, with the Rubikon being crossed. We suggest that it is a matter of ambidextrous leadership (formal, shared, or emergent) to foster a team’s decision-making in the pre-decisional ideation phase and, thereby, to lead the team across the Rubikon. Such a mindset change within the team from deliberation to implementation could be supported by goal orientation, autonomy, information absorption, information sharing, strategy and planning, as well as by coordination, as identified in our data. The mindset theory of action phases provides practical techniques for a mindset change such as formulating implementation intentions (Keller et al., 2020). Ambidextrous leaders could switch from opening to closing leadership behaviors by focusing on the formulation of implementation intentions and identifying the respective critical situations on the team level. In this way, they support the emergence of action-phase–related mindsets.
Results on our transit phase also provide guidelines for prioritization when organizations need to select team training contents due to limited budgets or time constraints: If prioritization is necessary, organizations and teams can focus on the categories in the transit phase. These not only support ideation and implementation but also facilitate the critical transit between the two. For example, training could focus on how to establish and use autonomy within teams, how to orient teams towards a shared goal, how to absorb and effectively share information, or how to plan and coordinate team efforts. By targeting training efforts on these shared influencing factors, teams could be more effectively equipped to enhance their innovative capabilities.
Limitations and a Future Research Agenda
Although the qualitative methodology allowed us to delve deeply into ideation and implementation in teams, our quantitative inferences rely on the authors’ coding based on participants’ statements. Due to possible biases when recalling events, these statements might merely reflect team members’ subjective theories about the success and failure of ideation and implementation in teams rather than objective mechanisms. Moreover, we examined single team members as representatives for their teams albeit studying a team level phenomenon (team innovation). However, the categories we gathered in this way also referred to team-level phenomena such as team diversity, team member characteristics, shared mental models, team spirit, participative safety, or information sharing (see Table 1). Although we examined team representatives (and not entire teams including all team members), we were thus able to gain insights into underlying team phenomena.
Nevertheless, future research should operationalize the categories we identified through existing and validated measures and longitudinally test their unique contributions to predicting ideation and implementation in whole teams in both successful and unsuccessful projects. This will allow research to quantitatively test whether our model can explain how the innovation process unfolds in successful innovation projects but also why some innovation projects may fail. In this way, also potential sampling biases could be restrained. Future data on ideation and implementation in whole teams, with additional information on team types (e.g., Hollenbeck et al., 2012) would allow to determine reliability measures of data (agreement of team members) and to assess potential differences in predictors of ideation and implementation in different types of teams.
An Agenda for Future Research on the Team Innovation Process.
Team Innovation as a Product of Effective Team Collaboration?
Classic frameworks on effective team collaboration differentiate between action processes and transition processes in teams (cf. Driskell et al., 2018; Mathieu et al., 2017, 2020). We did not explicitly focus on this distinction in our analysis, even though these classes of processes became especially evident in the implementation path and transit phase (e.g., coordination, monitoring vs. strategy and planning, information absorption). As we focused on team innovation rather than team effectiveness, we drew upon existing theorizing regarding team innovation (van Knippenberg, 2017). Nevertheless, future research might benefit from incorporating research on team effectiveness into theorizing on team innovation. In particular, the action regulation path of team innovation might benefit from effective teamwork. Farr et al. (2003) made such an endeavor by differentiating phases of action and transition within idea generation and idea implementation.
Team Innovation as Team Adaptation?
We also did not take a team adaptation perspective (cf. Georganta et al., 2020a, 2020b; Rosen et al., 2011) in our analysis, even though team innovation might resemble a team adaptation process given the highly unstable and dynamic nature of both processes. According to Burke et al. (2006), team innovation and team adaptation share core processes, and team innovation can be seen as a sub facet of team adaptation (see also Hülsheger et al., 2009).
Team Innovation, Team Adaptation and Effective Team Collaboration – Different but Same?
These content-related limitations bring us to a more comprehensive recommendation for future team research, that is, to display how team innovation processes, team adaptation processes and processes of effective teamwork can be disentangled or rather should be integrated within one comprehensive framework based on classic theories of human action and cognition due to their significant overlapping. Moreover, more complex and dynamic modern team structures are geared towards agility (cf. Driskell et al., 2018), so that integrating team innovation, team adaptation and team effectiveness becomes necessary from a practical stance as well.
In such an integrated framework, the multilevel structure of categories should also be considered. In our model, we did not specify the levels at which our categories are located; however, Bedwell et al. (2012) and Ramos et al. (2016) posit multilevel and cross-level effects, such as effects of an individual entity’s characteristics on the formation of collective entity phenomena or differential effects of categories located on different levels on idea generation and idea implementation (cf. Magadley & Birdi, 2012).
A prerequisite for such a comprehensive multilevel framework of team innovation, team adaptation, and team effectiveness would be a uniform definition of constructs and mapping onto higher-order categories. Salas et al. (2015) suggest clusters of categories relevant for teamwork, but team research requires agreement precisely on the exact definition of these categories’ content in terms of concrete constructs in order to conceptually align the field as a whole and thus make it more interpretable, integral, adaptive, and innovative.
Leadership and Team Innovation
Finally, many of the categories we identified are related (directly or indirectly) to (ambidextrous) leadership behaviors and leadership effects on the team’s environment. For example, leaders influence team innovation by making decisions on team composition in the hiring process and by creating a climate for creativity via role modelling (see participative safety in our model). Leaders formulate goals (see goal orientation), allocate resources (see e.g. management support) and make decisions on which ideas to pursue (see volition) (Hunter & Cushenbery, 2011). In this way, leaders can facilitate, direct and integrate creative efforts of team members (Mainemelis et al., 2015). Future research should thus incorporate a leadership perspective into the ImpAct model to show the criticality of essential leadership functions – fulfilled by a formally designated leader or within the collective work system – in the team innovation process.
Supplemental Material
Supplemental Material - From Ideation to Implementation: A Model of Team Innovation
Supplemental Material for From Ideation to Implementation: A Model of Team Innovation by Julia A. M. Reif, Carolin L. Feldmeier, Eleni Georganta, Katharina G. Kugler, and Felix C. Brodbeck in Group & Organization Management
Footnotes
Author’s Note
Some ideas described in this paper were presented at the 10th Convention of the Section for Economic, Work, and Organizational Psychology of the German Psychological Society, Dresden, Germany. Some of the data were collected as part of a master’s thesis completed by Carolin L. Feldmeier at the Universitaet Trier and as part of a bachelor’s thesis completed by Christina van de Ven at the Ludwig Maximilians-Universitaet Muenchen. We thank Keri Hartman for proofreading the manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was supported by the German Research Foundation (SFB 768).
Data Availability Statement
The datasets generated and analyzed during the current study are not publicly available due to the issues associated with anonymizing and deidentifying qualitative data but are available from the corresponding author on reasonable request.
Supplemental Material
Supplemental material for this article is available online.
Notes
Author Biographies
Associate Editor: Maria Simosi “AE”
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
