Why Team Performance Is the Wrong Thing to Focus On

Employees at the SpaceX Hawthorne office watching a launch test on October 13, 2024. At SpaceX, the focus is on the mission, not the team. Everyone feels part of the vision and the mission.

Many companies today are obsessed with teams.

The “old” approach of static departments and hierarchies is out. What’s in is teams – especially cross-cutting teams.

Amazon has defined their approach to this which they call “single-threaded teams”, which are teams that are set up to accomplish something and usually include people from many parts of the company. That’s a little different though, because Amazon’s single-threaded teams are often actually teams of teams.

Here we are talking about work that is innovative and cutting edge. Design and engineering work on cutting edge products must be innovative. It is not composed of repeatable processes. Rather, it is part craft, part creative, and relies a lot on experimenting – trying things to see if they work as you hope, and adjusting until you get it right and succeed.

Teams are often dwarfed by the fabric that links them together – the organization, and how the teams interact.

Today’s focus on teams probably originates from the groundbreaking work of Amy Edmondson, a Harvard professor (she was a PhD student at the time she began her work on teams) who discovered that organizations are most effective if they can create teams that have psychological safety, in the sense that team members are very willing to discuss outcomes, both good and bad, in the interest of improvement.

Dr. Edmondson’s work originated with teams in hospitals. That’s a challenging setting in which the stakes are literally life-and-death. However, the number of people needed to care for a patient is relatively small – a handful over time, if one counts all the specialists who provide a service, the lab techs, the nurses, and of course the doctors. In a lot of businesses, it often takes many more people to accomplish a goal – perhaps hundreds or thousands, organized as a great many teams. Consider for example the number of people involved in designing and developing a new car, or an airplane, or a new drug.

Thus, teams are often dwarfed by the fabric that links them together – the organization, and how the teams interact – just as human cells are dwarfed by the human body as the dominant system.

We have been taught to view team spirit as a good thing, perhaps because we have been raised to value sports teams and the team spirit associated with those. But in business, if people focus too much on their team, they lose sight of the organization’s overall goals: they become inward-looking and view their team as the boundary of their control and therefore their concern. That’s one of many reasons why sports teams are a very poor metaphor for teams in business.

I have seen dysfunction caused by excessive focus on one’s team many times. At one company where I consulted, software teams were able to complete a feature in days or at most two weeks, but then it took three months to integrate the feature with the work of other teams. Each team was very proud of their speed – they did not perceive a problem. But management lamented the “100-day cycle time” to get a feature to customers.

At the same company, there were teams who were responsible for maintaining what are called “core microservices”. Those microservices have “APIs”, which are specifications for how other software can access the microservices. A member of one of the core microservice teams told me that if another team uses the microservice team’s API incorrectly, then it is that other team’s problem, and not the concern of the microservice team. I responded, “But then the overall application will be broken”, and he shrugged his shoulders.

This behavior is easily explained by social identity theory, which describes how a team becomes, in effect, a tribe, and tends to view everything through the lens of what is best for the tribe, rather than for the organization as a whole. Teams also tend to create their own narratives, and resist forming consensus with other teams. In their book The New Psychology of Leadership, researchers Haslam, Reicher, and Platow write,

“people’s motivation to reach consensus, and their ability to do so, is structured by processes of self-categorization”.

[Ref: The New Psychology of Leadership: Identity, Influence, and Power, by S. Alexander Haslam, Stephen Reicher, and Michael Platow, 2011. Page 56.]

By “categorization” they mean selecting into a team, group, or tribe. In other words, separating ourselves by teams tends to divide us and cause us to resist agreeing with each other. That’s not what we want to have happening in an organization!

At highly agile companies like SpaceX (I was told by a senior manager there that they do not use the word “Agile”), there is not a focus on team spirit or team identity. Instead, there is a focus on the mission. It is common for (and I have anecdotes of this) teams to sacrifice their own progress in order to accommodate what another team needs, in the interest of success of the mission.

Teams are important. I don’t want to throw the baby out with the bathwater. Edmondson was on to something, and her work is brilliant. But I think the more important aspect of her work pertains to psychological safety and leadership, rather than specifically to teams. One can have psychological safety without teams, and if there is good leadership, good teams follow, and teams should come and go as needs change. There should be no heartache or sentimentality when a team disbands.

The more important aspect of Edmondson’s work pertains to psychological safety and leadership, rather than specifically to teams.

Teams are needed in order to decompose work and align people around a specific sub-goal of the overall mission goal. That is very powerful. But we need a balance with respect to our focus on the team or a focus on the overall mission. The latter should come first – the mission should be seen as far more important than the team or the team’s goals.

In later work, Edmondson shifted her focus to organizations. In a recent book called Teaming, Edmondson describes what she calls teaming (same as the book’s name). According to Edmondson, teaming is not about teams per se, but rather is about the behavior of people – their ability to quickly organize, have productive discussions, make decisions, and solve problems. They might not have time to even get to know each other, and as soon as the problem has been solved, the team disbands.

Notice that all of that is behavioral. It’s not a fixed team, and it’s not a workflow process. In fact, Edmondson even says,

“new employees instead are invited to get right to work helping to discover new processes.”

[Ref: Edmondson, Amy C.. Teaming: How Organizations Learn, Innovate, and Compete in the Knowledge Economy (p. 42). Wiley. Kindle Edition.]

In other words, the team members figure out how to organize things – they don’t follow an a priori process. And Edmondson says that is important, because it is through that self-construction of their methods that they learn.

As we say in our own leadership program, don’t give people answers – give them the right questions.

Edmondson refers to self-organizing teams but she is not talking about teams without leaders. When Edmondson refers to a self-organizing team, she is including all of the members of the team, including those who have been assigned leadership roles.

In work that is innovative and cutting edge, you will never have high-performing teams unless you have good leadership.

Recall that Edmondson’s early research was about medical teams. Certainly in the context of a surgical team that was assembled to perform a complex surgery, there might be several doctors of different specialties and there will be support professionals with specific roles. There is always a single doctor who is in charge, but that individual leads best through inclusive behavior, rather than the team having no lead and risking chaos (in her book, she discusses this).

When Edmondson uses the term “self-organization” (she uses the phrase twice in her book Teaming), she simply means that the team is not being told by the organization how it should be set up, such as that it should use Scrum or some other approach: the team decides, and the “team” includes its leader(s); and Edmondson goes into detail about how important it is for the team lead to use inclusive behavior, providing evidence from careful studies.

The elephant in the room is that none of this matters if the organization’s culture and the behavior of leaders is counter to what is needed. According to Edmondson (same source,),

“In the absence of a particular type of leadership, the right kinds of learning behaviors to deal with uncertainty and ambiguity tend not to occur. Teaming, it seems, requires a new type of leadership that supports speaking up, asking questions, and sharing ideas. In short, teaming requires a leadership mindset that cultivates an environment conducive to learning. I use the term “organizing to learn” to describe this leadership mindset and its accompanying practices.”

In other words, teams get their behavior from their leaders, and the culture of what is expected by leaders. Thus, to make teams perform well, one must make the organization’s leaders perform well – as leaders. One cannot turn teams into high-performing teams in isolation. Teams will become high-performing only if the organization’s leaders become good leaders.

That’s why obsession over teams is misguided. The focus should be on leadership. Good leaders generate psychological safety. Good leaders empower but also develop teams. Good leaders provide the support that is needed, but also make the right decisions. Good leaders create the right incentives that align everyone. Good leaders notice problems right away, because they ask questions, and they listen well when a situation is explained to them.

It’s all about leadership – not about teams. Leaders generate the behavior that results in what Edmondson calls “teaming”, through their expectations, coaching, mentorship, training programs they provide, structures, and incentives. Teaming – according to Edmondson – is entirely about behavior, rather than about teams per se. High-performing teams are an outcome of good leadership.

In work that is innovative and cutting edge, you will never have high-performing teams unless you have good leadership.