Why Teams Fail Even When People Are Capable

The customer had been promised delivery by Friday.

Sales sent the order to operations on Tuesday. Operations noticed that two specifications were missing but assumed sales would send them later. Finance had not approved the special price, but nobody knew who should follow it up. By Thursday afternoon, the manager was still telling senior leaders that everything was on schedule.

On Friday morning, the problem became impossible to hide.

Sales blamed operations for moving too slowly. Operations blamed sales for submitting an incomplete order. Finance explained that nobody had requested final approval. The manager called an emergency meeting, reassigned several tasks, and asked everyone to work through the weekend.

Every person involved was capable.

Each department had done something. Several people had worked hard. But the work failed while moving from one person to another.

This is where many teams fail: at the seams.

A seam is the point where information, responsibility, authority, or unfinished work must pass between people. One person stops. Another must continue. One department makes a promise. Another must fulfill it. A problem becomes visible. Someone must decide whether to raise it, hide it, solve it, or send it elsewhere.

Teams fail when they cannot reliably turn interdependent work into a shared result.

The failure may appear as a missed target, customer complaint, delayed decision, repeated mistake, exhausted manager, or another round of overtime. But the visible failure is usually the final consequence of a pattern the team has allowed to continue.

For a wider understanding of deliberate team practice, visit our complete guide to team building.

Capable people do not automatically create a capable team

Organizations sometimes respond to team failure by replacing a person.

They look for a more experienced manager, a stronger communicator, or an employee with a better attitude. Sometimes this is necessary. A person may lack the skill, integrity, or willingness the role requires.

But individual capability is only one part of team performance.

A talented employee can still fail inside a team with competing goals, confusing authority, hidden information, unsafe conversations, and incentives that reward people for protecting themselves.

Google reached a similar conclusion through Project Aristotle, its internal study of more than 200 teams. The researchers examined hundreds of possible factors, including team composition and individual performance. Their central finding was:

“Who is on a team matters much less than how team members interact, structure their work, and view their contributions.”

Google identified psychological safety, dependability, structure and clarity, meaning, and impact as the five dynamics that distinguished its more effective teams. The study did not prove that individual ability was unimportant. It showed that talent alone could not compensate for weak ways of working. Read Google’s account of the research.

J. Richard Hackman reached a related conclusion after decades of team research. In Leading Teams, he identified five conditions that prepare teams for strong performance: a real team, compelling direction, an enabling structure, a supportive organizational context, and competent coaching. His work shifts attention from trying to motivate people continually toward creating conditions that make effective teamwork more likely. See Hackman’s five conditions in Leading Teams.

When capable people repeatedly fail together, do not examine only the people.

Examine the game they have been asked to play.

Team failure usually begins before the missed target

A failed target is easy to see. The patterns that produced it are easier to ignore.

The project was late because the designer did not receive the final copy. The designer did not follow up because the project manager had reacted badly to earlier questions. The project manager believed everyone understood the deadline because it appeared on a slide. Team members agreed during the meeting but left with different interpretations of what “final” meant.

The deadline did not fail on the final day.

It had been failing quietly for two weeks.

This is why teams must learn to recognize early signals:

• The same decision keeps returning to the manager.
• People agree in meetings but resist afterward.
• Problems become visible only when deadlines are near.
• A few dependable employees repeatedly rescue the work.
• Departments complete their parts while the overall result remains unfinished.
• People send information but do not confirm that the next person can act on it.
• Nobody can explain clearly who owns the final move.
• After every failure, the team talks about effort instead of examining the pattern.

These are not minor annoyances. They are evidence that one or more of the team’s seams are beginning to break.

Four sources of team failure

The ten causes discussed below can be organized into four larger sources.

Direction failure happens when members are working toward different wins.

Design failure happens when the team’s roles, authority, membership, resources, or workflows make coordination difficult.

Interaction failure happens when people cannot exchange information, face friction, ask for support, and adjust their actions together.

Reinforcement failure happens when leaders and organizational systems reward the wrong behavior or allow harmful patterns to continue.

These sources often reinforce one another. Confusing direction creates role conflict. Role conflict creates frustration. Frustration becomes personal conflict. Leaders respond by taking control. Members become more dependent, and the manager becomes the team’s permanent rescuer.

The team does not have ten separate problems.

It has a connected pattern.

1. People are playing different games

A team cannot win together when members are protecting competing definitions of success.

Sales wants the largest order. Operations wants a workable delivery date. Finance wants a healthy margin. Customer service wants promises the company can keep. Each goal makes sense from inside the department.

The failure begins when people treat their functional target as the entire game.

Sales closes a deal that operations cannot deliver. Operations protects efficiency by refusing unusual customer requests. Finance delays a time-sensitive decision because the request does not follow the normal process. Every department can claim that it followed its responsibilities while the customer receives a poor result.

A compelling direction, in Hackman’s model, gives members a shared result that guides choices and trade-offs. Google’s research also found that effective teams understood their impact and how their work contributed to organizational goals.

At Team Bayanihan, we call this Play the Same Game.

A goal is not shared merely because it appears in the team presentation. It becomes shared when members use it to make consistent decisions—especially when protecting the shared win requires one department to adjust its preferred move.

Ask:

What result must all of us protect, even when our functional priorities pull us in different directions?

If members cannot answer that question in similar words, the team may already be playing several games.

2. The group is called a team, but its work is not designed as teamwork

Not every group of employees is a real team.

Some people report to the same manager but perform largely independent work. Others attend the same meetings but do not share an objective. Some project groups contain representatives from several departments, yet members remain accountable only to their home functions.

Putting people into one chat group does not create a team.

Hackman described a real team as having clear membership, reasonable stability, and interdependent work. Members must know who belongs to the team and why they need one another to produce the result.

Problems appear when team boundaries constantly change, essential contributors are excluded, or people are assigned without enough time and authority to participate. Members may be told to collaborate, yet their managers continue to prioritize unrelated departmental work.

The team exists on the organizational chart but not in the daily choices surrounding the work.

Before asking how the team should cooperate, establish whether a real team has been created:

• Is the membership clear?
• Do members share a result?
• Does the work genuinely require their contributions to connect?
• Do they have enough stability to develop useful working practices?
• Have they been given the authority and time to fulfill the assignment?

If the answer to these questions is no, another activity about cooperation will not repair the team’s design.

3. Roles and decision authority remain unclear

The most dangerous task in a team is often the one everybody believes someone else owns.

Several people may attend the discussion. Everyone may provide an opinion. But nobody leaves knowing who must make the decision, complete the action, or confirm that the result is finished.

Google’s Project Aristotle identified structure and clarity as one of the five dynamics of team effectiveness. Members needed clear roles, goals, plans, and decision processes.

Role clarity does not require a longer job description. It requires answers to practical questions:

• Who owns the next move?
• What contribution is expected from each member?
• Who makes the final decision?
• Whose input must be considered?
• What can be decided without the manager?
• When must the issue be escalated?
• How will the next person know that ownership has moved?

A team can have written roles and still experience ownership gaps. The breakdown often occurs between the roles.

Sales owns the customer relationship. Finance owns pricing approval. Operations owns delivery. But who owns confirming that all three have agreed before a customer promise is made?

The task does not disappear because it falls between job descriptions.

Strong teams make ownership visible at the seams.

4. Information is sent but does not become shared understanding

Many teams say they have a communication problem.

Often, they do not lack messages. They lack shared understanding.

Eduardo Salas, Dana Sims, and C. Shawn Burke reviewed teamwork research and proposed five core components: “team leadership, mutual performance monitoring, backup behavior, adaptability, and team orientation.” They also identified three coordinating mechanisms that connect those behaviors: shared mental models, mutual trust, and closed-loop communication. Read their “Big Five in Teamwork” research.

A shared mental model means that members hold a sufficiently similar understanding of the objective, situation, responsibilities, and plan. They do not need identical knowledge. They need enough common understanding to coordinate their different contributions.

Closed-loop communication means that information is not considered complete simply because somebody sent it. The receiver acknowledges it, confirms the meaning when necessary, and makes the next move clear.

Compare these two handovers.

The first says:

Please handle the customer request below.

The second says:

The customer needs 300 units delivered by Friday. The price is approved, but the color specification is still missing. Ana owns the customer clarification and will send it by 2 p.m. Operations can begin preparation now, but production must wait for Ana’s confirmation.

Both messages transfer information. Only the second helps the next person act.

At Team Bayanihan, we call this Serve the Next Person.

The sender’s responsibility does not end when the message leaves the outbox. It ends when the next person has what is needed to continue the work.

5. People learn that speaking early is dangerous

The manager says, “Tell me immediately when you see a problem.”

But when an employee raises a risk, the manager interrupts.

“Why are we hearing about this now?”

“Didn’t you think about that before?”

“Don’t bring me problems. Bring me solutions.”

The employee learns the real rule: speak only when you are certain, prepared, and able to solve the problem alone.

By then, the problem is expensive.

Amy Edmondson’s research describes psychological safety as a shared belief that a team is safe for interpersonal risk-taking. It allows people to ask questions, admit mistakes, request help, and raise concerns without expecting humiliation or punishment for speaking.

Psychological safety does not remove performance standards. It makes information about weak performance visible soon enough for the team to respond.

A 2023 review by Edmondson and Derrick Bransby examined 185 research papers on psychological safety. The research connects psychological safety with learning, knowledge-sharing, candor, inclusion, and team performance. Edmondson gives leaders a useful warning:

“If there’s no bad news, remind yourself: It’s not that it’s not there. It’s that you’re not hearing about it.”

Read the Harvard Business School discussion of the research.

When leaders punish the messenger, problems do not disappear.

They go underground.

6. Conflict becomes personal—or remains private

Some teams fail because people fight constantly. Others fail because nobody disagrees in the room where a decision can still be improved.

Both patterns are dangerous.

Frank de Wit, Lindred Greer, and Karen Jehn conducted a meta-analysis of 116 studies involving 8,880 groups. They found stable negative relationships between relationship conflict, process conflict, and group outcomes. The relationship between task conflict and performance was more complicated: disagreement about the work could help under some conditions, but its value weakened when it became entangled with personal and process conflict. Read the intragroup-conflict meta-analysis.

This research should make leaders careful about the popular statement that “conflict is good.”

Not all conflict is useful.

Task conflict asks, “Which approach gives us the stronger result?”

Relationship conflict says, “You are the problem.”

Process conflict argues repeatedly about who should do what, how resources should be used, or whose responsibility the work has become.

Teams also damage themselves when disagreement travels through private chats instead of the decision-making conversation. Members appear aligned during the meeting, then campaign against the decision afterward.

At Team Bayanihan, we call the needed commitment Face the Friction.

Members must learn to place relevant disagreement where it can help the shared result. They challenge the idea without attacking the person. They examine evidence. They clarify authority. After the decision is made, they support it unless important new information requires the team to reconsider.

The objective is not harmony.

The objective is honest coordination without unnecessary damage.

7. Teammates either abandon people or repeatedly rescue them

Interdependent work requires backup behavior.

Members notice when a teammate is overloaded, missing information, or facing a risk that may affect the shared result. They offer help before the failure spreads.

But teams frequently confuse support with rescue.

In abandonment, members say, “That is not my job,” while watching a preventable failure approach.

In rescue, the same dependable employees repeatedly complete other people’s commitments. The customer receives the result, but the team learns nothing. The original owner does not develop. The rescuer becomes exhausted. The manager sees a completed task and assumes the system works.

Healthy support keeps ownership clear.

A teammate can say:

I will help you solve this, but you remain responsible for closing the commitment.

At Team Bayanihan, this is Back Each Other Up.

Support should help people keep their commitments. It should not make ownership disappear.

8. The leader becomes the bottleneck—or leaves the team without leadership

Some managers become the center of every decision.

Members wait for approval, even when the decision falls within their expertise. Problems travel upward faster than authority travels downward. Capable employees stop exercising judgment because the safest move is to ask the boss.

When the manager is absent, the work slows down.

Other leaders make the opposite mistake. They tell members to take ownership but provide no direction, boundaries, priorities, or decision authority. The manager calls this empowerment. The team experiences abandonment.

Team leadership is one of the five components in the Salas teamwork model. Hackman also included competent coaching among the conditions supporting team performance.

Effective leaders do not make every move for the team. They create direction, clarify authority, protect standards, remove barriers, and help members improve their ability to coordinate.

Leadership must sometimes move closer to the work. The person with the most relevant information may need to guide the next move. But shared leadership does not mean unclear accountability.

Leaders cannot outsource teamwork to a facilitator. They influence what happens before, during, and after any team-building experience. This responsibility is explored further in the role of the team leader in team building.

9. The organization rewards the behavior it says it wants to stop

A chief executive tells departments to collaborate.

The performance system rewards each department only for its individual target.

Leaders tell employees to report risks early.

The people associated with bad news receive lower evaluations.

The company wants managers to develop people.

Managers are rewarded for immediate output, even when producing it requires them to take back delegated work.

Employees pay more attention to the scoreboard than the speech.

Hackman’s supportive-context condition includes access to information, education, resources, and rewards that strengthen team performance. When the surrounding system rewards selfish optimization, secrecy, internal competition, or heroic rescue, these behaviors will continue even after an inspiring workshop.

This does not mean people have no responsibility for their actions. It means leaders must examine what the organization repeatedly makes reasonable.

Ask:

• Which behavior do we publicly encourage?
• Which behavior do we actually reward?
• Who benefits when the present pattern continues?
• What risk does a person take by behaving differently?
• What system, permission, or consequence must change?

If collaboration produces a personal penalty, another speech about teamwork will not be enough.

10. The team completes the work but never examines how it worked

Teams move from one deadline to another.

A project finishes, and everyone immediately begins the next one. Mistakes are discussed only when somebody must be blamed. Success is celebrated without examining which decisions and practices made it possible.

The team gains experience but does not necessarily gain capability.

Adaptability is another component of the Salas teamwork model. Teams must notice changes, update their shared understanding, and adjust their behavior. That requires reflection.

Research supports the value of properly conducted debriefs. Scott Tannenbaum and Christopher Cerasoli’s meta-analysis found that individual and team performance could improve by approximately 20 to 25 percent when debriefs were conducted well. Read the debriefing meta-analysis.

A useful review does not ask only, “What went wrong?”

It asks:

• What result did we intend to produce?
• What actually happened?
• Where did our coordination help?
• Where did the work slow down or break?
• What did we assume that turned out to be wrong?
• What should we repeat?
• What must we change before the next attempt?
• Who owns the change?

At Team Bayanihan, Finish the Last Mile includes more than delivering the immediate output. It means closing commitments, capturing learning, and improving how the team will produce the next result.

The five seam failures

The Five Team Commitments provide a simple way to recognize where teamwork is breaking.

What keeps failing	The missing commitment	What should become visible
Members protect competing personal or departmental targets.	Play the Same Game	Decisions and trade-offs protect one shared result.
People are abandoned or repeatedly rescued.	Back Each Other Up	Members request and offer support while ownership remains clear.
Information and unfinished work break during handovers.	Serve the Next Person	The next person receives what is needed to continue immediately.
Concerns remain hidden or disagreements become personal.	Face the Friction	Members raise risks and differences while the team can still respond.
Actions, decisions, and promises remain unfinished.	Finish the Last Mile	Owners, deadlines, completion, and learning are made visible.

These commitments are not a complete replacement for good strategy, competent people, reasonable workloads, or supportive systems.

They help teams see what members must be able to expect from one another while pursuing the shared objective.

Do not diagnose the team from the final failure

When a deadline is missed, leaders usually begin with the person named on the task.

A better diagnosis traces the work backward.

Choose one important result the team failed to deliver reliably. Then examine the sequence:

1. What result did the team promise?
2. Where did the work first stop, slow down, or move backward?
3. Which people or departments were connected at that seam?
4. What did each person believe somebody else knew, owned, or would do?
5. What happened when the first concern appeared?
6. Which commitment was missing?
7. What leadership behavior or organizational system reinforced the pattern?
8. Where has the team repeated the same pattern before?

Suppose customer orders are frequently delayed.

“Poor communication” is not yet a useful diagnosis.

A closer examination may show that salespeople submit orders without complete specifications because their performance is measured only by booked revenue. Operations notices the missing information but waits because previous questions produced angry responses. Nobody owns confirming that the order is ready to enter production.

Now the team can see three connected causes: a broken incentive, avoided friction, and an ownership gap at the handover.

For a complete process, read How to Identify the Real Team Problem.

Replace one failing pattern with one practical commitment

Do not try to repair all ten causes during one meeting.

Begin with one important result and the seam that repeatedly prevents the team from reaching it.

Use this movement:

Objective → Friction → Commitment → Practice → Proof

Suppose the objective is to confirm every complete customer order within 24 hours.

The friction is an incomplete handover between sales and operations. Sales assumes operations will search through email threads for missing details. Operations waits for information but does not notify the owner immediately.

The team creates this commitment:

Before an order is handed to operations, sales will complete the five required order details. Operations will acknowledge receipt within two hours and identify any missing information immediately. The salesperson remains the owner of obtaining customer clarification.

The team practises this commitment on the next five live orders.

The proof may include fewer clarification messages, fewer orders returned to sales, faster confirmation, and a higher percentage of orders released within 24 hours.

This does not require a grand culture campaign.

It requires a visible commitment, repeated where the real work happens.

Can team building repair a failing team?

Team building can help when members need to see their patterns, practise coordination, face an important conversation, or make clearer commitments.

A well-designed experience may reveal how people respond to time pressure. One member takes control. Several people wait. Information stays inside one group. Nobody checks whether the next person understands the plan. The facilitator helps participants examine where the same pattern appears at work.

But team building has limits.

It cannot repair an impossible strategy, abusive leadership, missing resources, unqualified staffing, or incentive systems that punish the very behavior the organization wants. It cannot make people trust a leader who continues to retaliate against honest feedback.

Sometimes the team must change how it works.

Sometimes leaders must change the conditions surrounding the work.

Usually, both must move.

Before choosing an intervention, examine what team building can—and cannot—fix.

Stop asking why capable people cannot work together

The question assumes that capability belongs only to individuals.

Change the question.

Ask:

What does this team repeatedly make difficult, unsafe, unclear, or unrewarding?

Capable people can fail when they protect different wins. They can fail when responsibilities overlap but ownership disappears. They can fail when information is sent without becoming shared understanding. They can fail when speaking early is dangerous, conflict becomes personal, support becomes rescue, or every decision returns to the manager.

They can also succeed once the team changes the conditions and commitments shaping those moments.

Study the seams.

Find where the work stops moving. Name the friction without shaming the people inside it. Decide what members must be able to expect from one another. Practise the new commitment during real work, and look for proof that the result is moving.

Teams do not fail simply because people lack talent.

They fail when capable people are placed inside patterns that keep their capabilities from combining.

If your organization wants an experience built around the real seam your team needs to strengthen, explore our facilitated team-building workshops.

Books and research behind this article

• J. Richard Hackman, Leading Teams: Setting the Stage for Great Performances (2002).
• Eduardo Salas, Dana E. Sims, and C. Shawn Burke, “Is There a ‘Big Five’ in Teamwork?” Small Group Research (2005).
• Amy C. Edmondson, “Psychological Safety and Learning Behavior in Work Teams,” Administrative Science Quarterly (1999).
• Amy C. Edmondson and Derrick P. Bransby, research review on psychological safety (2023).
• Frank R. C. de Wit, Lindred L. Greer, and Karen A. Jehn, “The Paradox of Intragroup Conflict: A Meta-Analysis,” Journal of Applied Psychology (2012).
• Scott I. Tannenbaum and Christopher P. Cerasoli, “Do Team and Individual Debriefs Enhance Performance? A Meta-Analysis,” Human Factors (2013).
• Google re:Work, “Understand Team Effectiveness.”