In-Depth: Stable Or Fluid Teams? What Does The Science Say?
An investigation of three schools of thought on team development and high-performing teams
You can also listen to a podcast of this article.
Recently, the concept of “fluid teams”, “dynamic reteaming” or “ad-hoc teaming” has gained traction in the Agile community. Although the concept has many different definitions, a characteristic they share is that members move in and out of a team during its lifetime. There is obviously a wide range of fluidity, ranging from changes on a Sprint-by-Sprint basis to relatively infrequent changes. The notion of team fluidity challenges a commonly held belief that teams should strive to remain as stable as possible.
Despite this, decades of academic research into teams and workgroups have underscored the importance of team stability as a requirement for high performance. Although these studies did not compare stable teams versus fluid teams specifically, the most reliable theories we currently have to understand team development also seem to favor stability over fluidity.
I want to use this post to explore the research in this area. Considering just how popular the notion of fluid teams has become, I think it is important to weigh the evidence that supports it or contradicts it. This post is my attempt to do this.
This post is part of our “in-depth” series. Each post discusses scientific research that is relevant to our work with Scrum and Agile teams. With this series, we hope to contribute to more evidence-based conversations in our community and a stronger reliance on robust research over (only) personal opinions. We can use your support to fund the research and efforts that go into writing posts like these.
The need for fluid teams
The notion of fluid teams has been around for a long time, mostly in aviation, healthcare, and the armed forces. Bushe and Chu (2011) identify seven situations that drive the use of fluid teaming in those environments:
- High turnover among employees, leading to changes in teams
- Rapid downsizing or upsizing (e.g. crisis responses, armed conflict)
- A desire for different skills at different levels of the work that a team does.
- A desire for flexible allocation of personnel
- A desire to provide career development opportunities
- A desire to avoid toxic behavior by employees
- A desire to promote interaction and community in high-reliability workplaces.
The first two are imposed by the environment, whereas the other five are generally the result of decisions within the organization. Pryzbilla and his colleagues (2020) analyzed fluid teaming in IT environments and found that it is often used to bring highly specialized skills together to solve difficult problems. Edmondson (2012) argues that dynamic teaming is important to share and encourage learning. The professional literature around fluid teaming tends to conceptualize it as a necessary response to the reality that no team is actually stable. This concern is also present in academic literature, where authors like Mathieu and his colleagues (2008) and Mortensen and Haas (2018) point out that there is a research gap because studies are often based on stable teams, whereas this is often not the case in the modern workplace. This is also what inspired Edmondson (2012) in her academic work on “teaming”.
What defines fluid and stable teams?
So when is a team “stable” and when is it “fluid”? Unfortunately, both the professional and the academic literature offer many different definitions. But a common theme is that stable and fluid teams vary in the strength of their boundaries (Mortensen and Haas, 2018). Stable teams have very clear boundaries, where everyone knows who is in the team and who is not. The boundaries are less clear in fluid teams, where membership can change project-by-project or even task-by-task. The second theme is time. Teams are increasingly fluid as the frequency by which members join and leave changes over time.
No team is perfectly stable. All teams see changes in their membership after their formation, especially when they are long-lived. This is a natural consequence of changes in the workforce, personal careers, and life choices. I’ve worked with many teams that I would characterize as “stable”, yet all of them saw some changes in membership over the period of a year. It really is a matter of degree.
As we will see below, current scientific models for team formation underscore the need for time and frequent interaction to allow teams to develop the tissue that makes them high-performing. Even a single change in team membership can disrupt that process, and consequently, make it harder for teams to become high-performing sooner. But none of these models explicitly require teams to remain stable throughout their entire lifecycle. There may indeed be a bias towards stable teams in the research driving these models (Mortensen and Haas, 2018), but it is also possible that these models simply capture the social and cognitive reality of putting a group of people together into a team.
Below, I will discuss insights from three schools of thought on team development and formation, and what implications they have for team fluidity.
Insight #1: Team cognition and teams
Starting this millennium, a lot of research on team performance has focused on the notion of team cognition. One way to think about this is as a “distributed team mind”, with its own memory and mental models, that are shared across its members. Examples of such mental models are “who in this team has certain skills?”, “what information do different members need to perform their tasks well?” and “how do we effectively combine our skills to work together effectively”. Research has linked team cognition to higher performance and motivation (Mathieu et al, 2000), increased effectiveness (Kearny, Gebert & Voelpel, 2009), and generally explains a substantial amount of the variance (~19%) in the effectiveness of teams (De Church & Mesmer-Magnus, 2010).
A good analogy is to understand team cognition as muscle memory. When you try to ride a bicycle for the first time, it takes great effort to stay upright and move forward. Through repetition, your body develops the reflexes to do this automatically — outside of your immediate awareness. Team cognition is similar in the sense that the entire team can be thought of as a body, where each member (the limbs) has to learn how to coordinate their work effectively to move forward. This shared understanding is captured in the “team mental models” that drive team cognition.
Just like muscle memory, team cognition requires time to form as the members of a team work together on shared tasks. In turn, this facilitates further collaboration. Kozlowski & Ilgen (2006) describe this reciprocity as “process begets structure, which in turn guides process”.
Even small changes to a team can fracture team cognition by scooping out part of the “team mind”. The team has to effectively rebuild the mental models about who knows what in a team, as well as how to collaborate on tasks in this new configuration. This isn’t just a matter of creating a new skill matrix but also rebuilding habits (mental models) for how decisions are made in a team, how conflict is navigated, how goals are set, and how problems are addressed collectively.
Unfortunately, I haven’t found research on team cognition that specifically investigated just how much time is needed to develop such mental models. But this is bound to be deeply contextual, and dependant on the complexity of the tasks, the skills of members, and the effort invested in specifically developing such models.
Learn more about team cognition in this extensive post I wrote earlier.
Insight #2: Cohesion and small teams
Another aspect of teams is how cohesive they are. Cohesion — or social cohesion, esprit de corps, commitment — has been studied extensively by social psychologists. A group is considered cohesive when its members are attracted by the idea of their group (Hogg, 1992). That is; the members like the idea of being part of their group and they self-identify as members of it. We see cohesive groups everywhere in life. From sports fans who unite behind their team to religious communities. And a group of friends, or a family, can also be more or less cohesive. But we obviously also find cohesive groups in the workplace. Organizational psychologists found that motivation is higher in cohesive groups (Beal et. al., 2003). This effect is stronger for smaller teams than larger teams. They also tend to perform better (Evans & Dion, 2012), are better able to deal with stress and pressure (Salas, Driskell & Hughes, 1996). Wang et al (2006) studied software teams tasked with ERP implementations and found that cohesive teams performed significantly better than less-cohesive teams.
Very few studies have investigated how teams become cohesive over time, or how much time is required in order to improve performance. However, several studies show that cohesion positively impacts performance only in later stages of team development (e.g. Bradley et. al., 2013). Kozlowski and Ilgen (2006) reviewed existing literature on the effectiveness of teams and workgroups and conclude that “there is solid research-based evidence for the importance of cognitive (unit-team climate, team mental models, and transactive memory), motivational (team cohesion, team efficacy, and potency), and behavioral (team competencies, functions, and regulatory mechanisms) processes and emergent states”. Although these studies don’t answer the question of how much time is needed, they all emphasize that such team-level processes — like cohesion, team memory, and team climate — emerge over time.
Read more about cohesion in this extensive post I wrote earlier.
Insight #3: Team formation stages
Organizational and social psychologists have long studied how teams develop to become high-performing. The most commonly known model is Tuckman’s Stages of Group Development (1965). Based on research with therapy groups, Tuckman argued that most teams move through four discrete stages in their formation. The first stage is forming, where members get to know each other and get clear on the tasks at hand. The second stage is storming, where members begin to navigate their different individual preferences for how to perform work, what quality to expect and how to interact. The conflict in this stage can be subtle or overt. When the members manage to successfully navigate the conflicts from this stage, they move into the norming stage. This is where members commit to the shared norms about work, quality, and interaction. With this groundwork in place, teams now move into the performing phase and become high-performing.
Tuckman’s model remains one of the simplest and easiest to explain models for team development to this day. Despite its popularity, the actual empirical evidence for the stages or the order is minimal (Bonebright, 2009). Most of the academic criticism has focused on the assumption that all teams go through them in the same order.
Since Tuckman, many models for team development emerged that don’t assume a linear progression through stages. For example, Mcgrath's (1991) Time-Interaction-Performance (TIP) Theory simply identifies four modes — inception, technical problem solving, conflict resolution, and execution — that teams generally go through, but in different and complex orders. There is also the Dynamic Organic Transformation (DOT) team model for high-performance teams (Courtney, Navarro & O’Hare, 2007) or Kozlowski’s process model for team development (1999).
Again, there is no research that investigates specifically how much time is required. Or if these processes can be fast-tracked. This is probably due to the difficulty of measuring many teams over a long period of time.
What all these models have in common is that they treat team formation as a process that teams go through in order to become high-performing. As members interact and spend time together, they learn how to effectively coordinate their individual work to reach shared objectives, and improve their performance as a team over time. This process can also be disrupted or hindered by changes in teams or their environment.
Four key findings
We explored insights from three schools of thought on team development and formation. These are the key findings:
- Two ingredients that connect all three insights are time and interaction. It takes time and frequent interaction for teams to develop the muscle memory they need to become high-performing. Changes to teams can disrupt this process.
- Theories around team cognition and team cohesion provide a good explanation for why teams move through different stages of development.
- Much of the academic research to date has focused on stable teams, where membership remains more or less the same over their lifecycles. But teams that are stable to that extent are rare in modern workplaces.
- There is very little research that investigates specifically how much time it takes for teams to become high-performing, and whether or not such processes can be fast-tracked through training, coaching, and other interventions.
A quick note on who initiates membership changes
The academic literature on teams is still largely based on organizations that are typically manager-led. This means that many papers on fluid and stable teams assume that membership changes are initiated by managers or HR, possibly on behalf of members themselves. I would like to emphasize that in my treatment of this subject, I assume that team changes are mostly initiated by the teams themselves. This corresponds with research on autonomous teams, as well as principles of Agile software development.
A quick note on a potential false dichotomy
Some readers have argued that the dichotomy between stable and fluid teams is a false one. In their understanding, fluid team designs propose a stable pool from which subteams are dynamically drawn for the tasks at hand. Pools consist of a dozen to many dozen members, whereas subteams are much smaller. In this view, teams are both flexible (subteams) and stable (pools).
I don’t see how this changes the reading of available research though. It merely moves the goalposts of what a “team” is for the sake of the argument. Academic studies on teaming generally define teams as groups of 3–8 people that work on a shared task. So in the context of fluid teaming, we’re more interested in the subteams, and these do change frequently. As I outlined in this post, we have very strong theoretical and empirical reasons to assume that this will negatively impact their performance. However, fluid team approaches assume that such negative impacts won’t be as severe, or even non-existent if the subteams are at least drawn from a larger stable pool. While I wish that were true, it is clearly an untested hypothesis that has the weight of existing scientific evidence against it.
So … stable teams or fluid teams?
The research we explored does not provide a clear answer as to whether stable teams are better or worse than fluid teams. Such a question would require us to follow both stable and fluid teams over a period of time and compare their performance. It is unlikely that research of this kind will happen any time soon, as it is very complex and time-involving.
However, the studies and the (empirically grounded) theories we explored all suggest that teams benefit from time and interaction if their aim is to become high-performing. The processes that happen between members during this time and these interactions can be disrupted and impeded by changes in the team. Theoretical models like team cognition and team cohesion provide a good explanation as to why this is so. And there is ample empirical evidence from actual teams to support these models.
“However, the studies and the theoretical models we explored all suggest that teams benefit from time and interaction if their aim is to become high-performing.”
If there is anything we should take from these models is that we should be careful with teams. When you put a group of people together, they start to form a social fabric of norms, expectations, and knowledge between each other. Every time the membership of a team changes, part of this fabric has to be rewoven in order to regain the same level of performance. And that takes time and conscious effort.
This does not mean that fluid teams are a bad idea. No team is perfectly stable. It is a matter of degree. There are certain scenarios where there is no option but to change the membership of a team per project, or even more frequently than that. And if team members want to leave, they obviously should always be able to. But based on the findings outlined in this post, organizations should probably not expect the same level of performance from fluid teams. However, if organizations are serious about fluid teams they do well to follow lessons from organizations that already have ample experience with fluid teams; like teams in aviation, healthcare, and the armed forces. These organizations invest deeply in shared training, skill standardization, and extensive support networks to make sure that teams can perform well, regardless of who makes up the teams (Bushe & Chu, 2011).
The wish behind fluid teaming risks becoming the father of its thought. For many managers in modern organizations, the notion of team fluidity could be seen as an excuse to avoid deep investments in stable teams and the skills of their members. Or to downsize teams. Or to hire cheaper employees to replace more expensive ones. It could be chosen simply out of convenience. I know that every advocate of team fluidity would object strongly to such applications. But most practitioners who work with teams on a day-to-day basis can easily come up with situations where managers cherry-picked only the parts they liked and ignore the rest.
The recommendation based on the evidence is one of caution. Yes, fluid teaming has its place in team design practices. But without stronger empirical evidence that compares fluid teaming practices with more stable teams, it still seems prudent to aim for stability over fluidity. For those organizations that want to experiment with fluid teaming, it is wise to assess their capabilities to support, train and develop stable teams. The support structure that is necessary there will be even more important as teams become more fluid. If organizations are unable to develop high-performing teams from mostly stable teams, it is very unlikely they will succeed with more fluid team designs.
“ If organizations are unable to develop high-performing teams from mostly stable teams, it is very unlikely they will succeed with more fluid team designs.”
The concept of team fluidity challenges preconceived notions about teams. It also challenges the assumption that teams are closed systems with clear boundaries. But every practitioner who works with, or as part of, teams already knows that. Teams are rarely stable in modern workplaces. The key question is: should we also encourage teams to be flexible in their memberships, or should we strive towards stability where possible?
I will leave the closing words with Bushe & Chu (2011), who reviewed existing research on fluid teams: “Fluid teams may never have the same potential as stable teams to develop into synergistic, high performing teams, but sometimes the situation makes them unavoidable”.
This highlights that fluid teams certainly have their place in team design. But it is more of a last resort if it is impossible to strive for stability. The current academic literature in this area favors more-or-less stable teams over more fluid teams, as do the models we have to understand what makes teams perform better than others. More empirical research with sufficiently large samples and proper designs would be very useful to offer more guidance to practicioners.
Beal, D. J., Cohen, R. R., Burke, M. J., & McLendon, C. L. (2003). Cohesion and performance in groups: a meta-analytic clarification of construct relations. Journal of applied psychology, 88(6), 989.
Bonebright, D. A. (2010). 40 years of storming: a historical review of Tuckman’s model of small group development. Human Resource Development International, 13(1), 111–120.
Bradley, B. H., Baur, J. E., Banford, C. G., & Postlethwaite, B. E. (2013). Team players and collective performance: How agreeableness affects team performance over time. Small Group Research, 44(6), 680–711.
Brown, R., & Pehrson, S. (2019). Group processes: Dynamics within and between groups. John Wiley & Sons.
Bushe, G. R., & Chu, A. (2011). Fluid teams: solutions to the problems of unstable team membership. Organizational Dynamics, 40(3), 181–188.
Cannon‐Bowers, J. A., & Salas, E. (2001). Reflections on shared cognition. Journal of Organizational Behavior: The International Journal of Industrial, Occupational and Organizational Psychology and Behavior, 22(2), 195–202.
Comer, D. R. (1995). A model of social loafing in real work groups. Human relations, 48(6), 647–667.
Courtney, H. S., Navarro, E., & O’Hare, C. A. (2007). The Dynamic Organic Transformational (DOT) team model for high‐performance knowledge‐worker teams. Team Performance Management: An International Journal.
DeChurch, L. A., & Mesmer-Magnus, J. R. (2010). The cognitive underpinnings of effective teamwork: a meta-analysis. Journal of applied psychology, 95(1), 32.
Evans, C. R., & Dion, K. L. (2012). Group cohesion and performance: A meta-analysis. Small Group Research, 43(6), 690–701.
Forsyth, D. R. (2009). Group Dynamics, 5th. Belmont, CA: Wadsworth.
Hogg, M. A. (1996). Intragroup processes, group structure and social identity. Social groups and identities: Developing the legacy of Henri Tajfel, 65, 93.
Kearney, E., Gebert, D., & Voelpel, S. C. (2009). When and how diversity benefits teams: The importance of team members’ need for cognition. Academy of Management journal, 52(3), 581–598.
Kozlowski, S. W. J., Gully, S. M., Nason, E. R., & Smith, E. M. (1999). The changing nature of work performance: Implications for staffing, personnel actions, and development. SIOP Frontiers Series, 240–292.
Kozlowski, S. W., & Ilgen, D. R. (2006). Enhancing the effectiveness of work groups and teams. Psychological science in the public interest, 7(3), 77–124.
Przybilla, L., Wiesche, M., & Thatcher, J. B. (2020). Conceptualizing Fluid Team Membership and Its Effects in IT Projects: A Preliminary Model.
Mathieu, J. E., Heffner, T. S., Goodwin, G. F., Salas, E., & Cannon-Bowers, J. A. (2000). The influence of shared mental models on team process and performance. Journal of applied psychology, 85(2), 273.
Mathieu, J., Maynard, M. T., Rapp, T., & Gilson, L. (2008). Team effectiveness 1997–2007: A review of recent advancements and a glimpse into the future. Journal of management, 34(3), 410–476.
McGrath, J. E. (1991). Time, interaction, and performance (TIP) A Theory of Groups. Small group research, 22(2), 147–174.
Mortensen, M., & Haas, M. R. (2018). Perspective — Rethinking teams: From bounded membership to dynamic participation. Organization Science, 29(2), 341–355.
Tuckman, B. W. (1965). Developmental sequence in small groups. Psychological bulletin, 63(6), 384.
Wang, E. T., Ying, T. C., Jiang, J. J., & Klein, G. (2006). Group cohesion in organizational innovation: An empirical examination of ERP implementation. Information and Software Technology, 48(4), 235–244.
Whitworth, E., & Biddle, R. (2007, August). The social nature of agile teams. In Agile 2007 (AGILE 2007) (pp. 26–36). IEEE.