Five dysfunctions of a team

Patrick Lencioni's five dysfunctions model, from The Five Dysfunctions of a Team: A Leadership Fable (2002): what each dysfunction looks like, how to diagnose it, the leader's job and the book's exercises for fixing it, plus the evidence-based team research that complements it (psychological safety, Project Aristotle, Tuckman, Hackman). The team leader's personal stance is in extreme ownership, intent and delegation in mission command, one-on-ones, feedback and hiring in people management, and how teams decide in decision-making models.

The model at a glance

The book is a novel about Kathryn Petersen, the new CEO of a fictional start-up, DecisionTech, fixing an executive team that is smart and well paid but not working as a team. The last part of the book sets out the model directly.

Why a pyramid

Lencioni's argument is a causal chain, read from the bottom:

• Without trust, people won't risk open disagreement, so conflict goes underground.
• Without real conflict, people haven't aired their views, so they don't truly buy in to decisions.
• Without commitment, nobody has signed up to a clear plan, so nobody feels entitled to hold a peer to it.
• Without accountability, standards slip and people drift to what earns them personal status.
• The top dysfunction is the one that actually costs you: the team stops focusing on results.

The practical conclusion: you can't fix accountability with a new dashboard if people don't trust each other enough to argue. Work from the base upward, while keeping the results at the top in view.

Trust: vulnerability-based vs predictive

Lencioni defines team trust as confidence among members that their peers' intentions are good, so there is no reason to be protective or careful around the group (paraphrased). He distinguishes it from the everyday kind.

A team can have lots of predictive trust (everyone knows who will be defensive) and none of the vulnerability kind.

1. Absence of trust

Shows up as invulnerability: people hide mistakes and weaknesses, don't ask for help, and assume bad intent.

Diagnostic questions

• When did someone on this team last say "I was wrong" in front of the others? Who was it?
• Does anyone ask a peer (not the boss) for help with something in their own area?
• Do problems surface in the meeting, or in the corridor and in DMs after it?
• If a teammate missed a deadline, would the first assumption be incompetence, laziness, or a reason we don't know?

Leader's role: demonstrate vulnerability first, genuinely (not staged), and create an environment that does not punish it. If the leader never admits a mistake, nobody else will.

2. Fear of conflict

Shows up as artificial harmony: meetings are polite and dull, real disagreements are voiced only afterward, and important topics are avoided. Lencioni means productive ideological conflict about concepts and ideas, not personal attacks. He places it on a continuum from artificial harmony at one end to mean-spirited personal attacks at the other, with the ideal point near the middle; most teams sit too far toward harmony.

Diagnostic questions

• When did the team last change its mind in a meeting because of an argument?
• What's the topic everyone knows about but nobody raises?
• Are the real decisions made in the meeting, or in one-to-ones and hallways afterward?
• After a meeting, do people say to each other what they didn't say during it?

Leader's role: show restraint when people engage in conflict, and let resolution happen naturally rather than stepping in early to protect people. Model healthy conflict yourself: disagree with a peer openly and respectfully.

CONFLICT NORMS (agree as a team, revisit quarterly)
1. We argue about ideas, never about people's motives.
2. Silence is not agreement. The chair asks each quiet
   person for their view before a decision.
3. Bad news goes first on the agenda.
4. "I disagree because…" is always welcome; sarcasm isn't.
5. Anyone can call "real-time permission" when a debate
   gets uncomfortable but is still useful.
6. We don't relitigate a decision in hallways. Reopen it
   only in the room, with new information.
7. Heat is fine; contempt, eye-rolling and interruptions
   aren't. Anyone can call it out.

3. Lack of commitment

Shows up as ambiguity: decisions are vague, people leave meetings with different understandings, and work stalls while people wait for more data or for consensus. Lencioni names two causes: the desire for consensus and the need for certainty. His point is that reasonable people don't need their way to buy in; they need to have been heard and to know their view was considered. Amazon's version of the same idea is "disagree and commit" (Bezos, 2016 shareholder letter).

Diagnostic questions

• At the end of the last meeting, could everyone write down the same three decisions?
• What did different team members tell their own reports about that meeting?
• How many decisions from last month were reopened without new information?
• Who has the final call when the team can't agree, and does everyone know it?

Leader's role: be comfortable making a decision that ultimately turns out to be wrong, keep pushing for closure, and hold the team to the schedules it sets. Don't let the team hold out for certainty or unanimity.

CASCADING COMMUNICATION (last 10 minutes of every meeting)
Decision 1: ______________________________________
  Owner: ________  By: ________
  What we tell our teams (same words): ____________
  Confidential until: ____________
Decision 2: ...
Check: does anyone's version differ? Fix it now.
Send: each member tells their team within 24 hours.

4. Avoidance of accountability

Shows up as low standards: people won't call peers on behavior or performance that hurts the team, so mediocrity spreads and the leader becomes the only source of discipline. Lencioni means peer-to-peer accountability, which he argues is more effective than accountability that runs only through the boss. It depends on commitment: you can't hold someone to a plan they never signed up to.

Diagnostic questions

• When a peer's work slips, who says something first: a peer, the manager, or nobody?
• Are the team's goals and each person's commitments written down where everyone can see them?
• Is there anyone whose behavior everyone complains about privately and nobody addresses?
• Have standards drifted downward over the last six months without anyone deciding they should?

Leader's role: encourage and allow the team to be the first and primary accountability mechanism. When the team fails to act, the leader must be willing to be the ultimate arbiter of discipline, which is what keeps peer accountability credible.

TEAM COMMITMENTS BOARD (visible to everyone, reviewed weekly)
| Commitment              | Owner | Due    | Status | Note |
| Launch pricing page v2  | Ana   | 12 Oct | amber  | ...  |
| Hire 1 backend engineer | Raj   | 31 Oct | green  | ...  |
| Cut p95 latency < 300ms | Wei   | 30 Nov | red    | ...  |
 
Behavior standards: we reply to blocking questions the
same day; we tell a peer directly before escalating; we
flag slips as soon as we know, not on the due date.

5. Inattention to results

Shows up as status and ego: members care more about their own career, department or visibility than about the team's collective goals. Lencioni singles out team status (just being on the team is enough) and individual status (looking good personally) as the two distractions.

Diagnostic questions

• Can every member name the team's top three results for this quarter, and the current score?
• If the team hit its goals but your department missed one of its own, would you count it a good quarter?
• Do people argue harder for their own budget and headcount than for the shared goal?
• What gets celebrated: individual wins or team wins?

Leader's role: set the tone that results come first, be selfless and objective, and reserve rewards and recognition for people who make real contributions to the group's goals.

The first team

Lencioni's later book The Advantage (2012) names an idea already present in the fable: for every member of a leadership team, that leadership team is their "first team", with priority over the department they run. A VP of engineering is first a member of the executive team and only second the head of engineering.

For a small start-up, the "leadership team" may be the founders plus two or three leads. The same rule applies: the founding group's shared goals come before any one person's area.

Team self-assessment

A self-scoring checklist in the spirit of the book's 15-item assessment (paraphrased and extended, not the book's wording). Each member scores each statement independently: 3 = usually, 2 = sometimes, 1 = rarely. Add the three scores for each dysfunction.

TRUST
 [ ] T1 People admit their mistakes and weaknesses to the team.
 [ ] T2 People ask teammates for help without worrying
        how it looks.
 [ ] T3 People know about each other's lives and work
        and use it to understand each other.
CONFLICT
 [ ] C1 Discussion of important issues is passionate and
        unguarded.
 [ ] C2 Meetings tackle the most important and difficult
        topics rather than the easy ones.
 [ ] C3 People voice their real opinions even when they
        risk disagreeing with the leader.
COMMITMENT
 [ ] M1 Meetings end with clear, specific decisions that
        everyone can repeat.
 [ ] M2 People support decisions they argued against.
 [ ] M3 Decisions are not reopened without new information.
ACCOUNTABILITY
 [ ] A1 People point out unproductive behavior in peers.
 [ ] A2 Peers, not only the leader, raise performance
        slips.
 [ ] A3 People apologize to one another and hold each other
        to what was agreed.
RESULTS
 [ ] R1 Morale is hit by missing team goals, not only
        personal ones.
 [ ] R2 People give up credit and resources for the team's
        results.
 [ ] R3 People put team goals ahead of their department's.
 
SCORING (per dysfunction, 3 statements, 3-9 points)
 8-9  probably not a problem
 6-7  could become a problem
 3-5  probably needs addressing now
Compare individual scores: a wide spread on one area is
itself a finding.

The book's scoring bands are the same as those above. Lencioni's publisher also sells a commercial assessment, The Five Behaviors of a Cohesive Team (Wiley), built on the model.

Worked scenarios

What the evidence says: complementary models

Lencioni's model is practitioner wisdom. These models come from research and mostly point the same way.

Psychological safety (Edmondson)

Amy Edmondson defined team psychological safety as "a shared belief held by members of a team that the team is safe for interpersonal risk taking" (Edmondson 1999, Administrative Science Quarterly). In a study of 51 work teams in a manufacturing company, psychological safety was associated with team learning behavior, which in turn was associated with performance. The Fearless Organization (2018) is the practitioner book.

• It's close to Lencioni's vulnerability-based trust, but defined at the team level and measured with a validated survey.
• A meta-analysis of 136 samples (Frazier et al. 2017, Personnel Psychology) linked it to task performance and organisational citizenship behavior, among other outcomes.
• Edmondson stresses that safety isn't the same as niceness or low standards. She crosses safety with performance standards to get four zones:

Google's Project Aristotle

Google People Operations studied what made its own teams effective: over two years, 200+ interviews, 250+ attributes, 180+ teams (Julia Rozovsky, re:Work blog, 17 November 2015). Who was on the team mattered less than how members interacted, structured their work and viewed their contributions. Five dynamics:

Psychological safety was by far the most important, the base for the other four. Evidence quality: an internal study, not peer reviewed, with no published data. It is consistent with the academic literature, but don't cite it as if it were a controlled experiment.

Tuckman's stages

Bruce Tuckman (1965, Psychological Bulletin) reviewed studies of small groups, mostly therapy and training groups, and proposed forming → storming → norming → performing. Tuckman and Mary Ann Jensen added adjourning in 1977.

Caveats: it's a descriptive model, not tested as a fixed sequence; teams can loop back (a new member can restart storming). Connie Gersick's study of project teams (1988, Academy of Management Journal) found a different pattern: inertia until roughly the midpoint of the deadline, then a burst of change. That suggests deadlines matter as much as stages.

Hackman's conditions for team effectiveness

J. Richard Hackman (Leading Teams, 2002) argued that a leader can't make a team great; they can only create conditions that make greatness more likely.

• In the 2009 HBR interview "Why Teams Don't Work", Hackman reported that across more than 120 senior leadership teams, fewer than 10% of members agreed about who was on their own team.
• He also cited evidence that newness is a liability: an NTSB finding that 73% of incidents in its database happened on a crew's first day flying together.
• Hackman and Ruth Wageman's rough 60–30–10 rule: about 60% of the difference in team performance comes from how the team was designed, 30% from how it was launched, and 10% from coaching once work is under way. It's a rule of thumb from their research, not a precise estimate.

Team size

Hackman: "My rule of thumb is no double digits" (HBR 2009); he capped student teams at six. Links between members grow quadratically:

Amazon's two-pizza team rule, attributed to Jeff Bezos, says the same thing: a team should be small enough to feed with two pizzas. The exact number matters less than keeping teams small enough for everyone to know what everyone else is doing.

Lencioni vs Edmondson vs Aristotle vs Hackman

The models fit together: Hackman tells you how to set up a team, Edmondson and Aristotle what climate it needs, Lencioni what behaviors to look for and practice inside it.

Criticisms and evidence quality

Plainly: the five dysfunctions are a memorable, practically useful checklist that lines up well with psychological safety research, but the pyramid is not an empirically validated model. Use it to start conversations and structure interventions; use Edmondson's survey or Aristotle's five questions if you want to measure.

A 90-day plan to fix a dysfunctional team

1. Days 1–14: diagnose

   • One-to-one with every member: what's working, what's not, what they'd change, what they haven't said in meetings.
   • Everyone completes the self-assessment independently; you tabulate averages and spread per dysfunction.
   • Check Hackman's conditions: is the team real, small enough, with a compelling direction and the resources it needs? Fix structural problems first; no exercise will overcome a badly designed team.
   • Write one page: the top two dysfunctions, with evidence.

2. Days 15–30: launch or relaunch (trust)

   • A half-day or full-day off-site. Share the assessment results openly.
   • Leader goes first: a real mistake, a real weakness, a request for help.
   • Personal histories exercise; if trust allows, the team effectiveness exercise.
   • Agree conflict norms and a short list of behavior standards.

3. Days 31–60: conflict and commitment

   • In each weekly meeting, put the hardest topic first. Assign a rotating "miner of conflict".
   • Use real-time permission whenever a debate gets uncomfortable but useful.
   • End every meeting with cascading communication: decisions, owners, dates, what we tell our teams.
   • Use deadlines and worst-case analysis to close decisions that have been hanging.

4. Days 61–90: accountability and results

   • Publish team goals, owners and standards where everyone can see them.
   • Weekly progress review against commitments; peers give feedback before the leader does.
   • A team scoreboard of 3–6 results, reviewed at every meeting.
   • Move some recognition (and, where you can, rewards) to team outcomes.
   • Deal with anyone who persistently undermines the team; a team watches how the leader handles its worst behavior.

5. Day 90: re-measure

   • Rerun the self-assessment and compare. Look for movement in the lowest area and a narrower spread.
   • Keep what worked as permanent rituals; drop what felt like theatre.
   • Schedule a quarterly check-in to repeat the assessment.

QUARTERLY TEAM HEALTH CHECK (30 minutes)
1. Each person scores the five Aristotle questions 1-5:
   safety, dependability, clarity, meaning, impact.
2. Share scores simultaneously (avoid anchoring).
3. Discuss the lowest average and the widest spread.
4. Agree one norm or ritual to change. Owner + date.
5. Review last quarter's change: keep, adjust or drop.

Common mistakes when fixing a team

• Running trust exercises while ignoring a structural problem (unclear goals, wrong people, too big a team).
• Leader asks for vulnerability but never shows it, or reacts badly the first time someone admits a mistake.
• Treating conflict norms as a poster rather than enforcing them in the next heated meeting.
• Confusing consensus with commitment, and waiting for everyone to agree before deciding.
• Leaving accountability entirely to the leader, which teaches peers to complain upward instead of speaking directly.
• Measuring individual output only, then wondering why nobody attends to team results.
• Declaring victory after the off-site. Behavior change shows in the next 20 meetings, not the first.

References

• Patrick Lencioni, The Five Dysfunctions of a Team: A Leadership Fable (Jossey-Bass, 2002): the fable, model, exercises and 15-item assessment
• Patrick Lencioni, Overcoming the Five Dysfunctions of a Team: A Field Guide (Jossey-Bass, 2005): practical tools and exercises
• Patrick Lencioni, The Advantage: Why Organizational Health Trumps Everything Else in Business (Jossey-Bass, 2012): the "first team" and organisational health
• Edmondson (1999), Psychological Safety and Learning Behavior in Work Teams (opens in a new tab), Administrative Science Quarterly 44(2): the original construct and 51-team study
• Amy C. Edmondson, The Fearless Organization (Wiley, 2018): psychological safety for practitioners
• Frazier et al. (2017), Psychological Safety: A Meta-Analytic Review and Extension (opens in a new tab), Personnel Psychology 70(1): 136 samples
• Rozovsky (2015), The five keys to a successful Google team (opens in a new tab), Google re:Work: Project Aristotle's findings
• Tuckman (1965), Developmental sequence in small groups (opens in a new tab), Psychological Bulletin 63(6): forming, storming, norming, performing
• Tuckman and Jensen (1977), Stages of Small-Group Development Revisited (opens in a new tab), Group & Organization Studies 2(4): adds adjourning
• Connie J. G. Gersick, "Time and Transition in Work Teams", Academy of Management Journal 31(1), 1988: the midpoint-transition finding
• J. Richard Hackman, Leading Teams: Setting the Stage for Great Performances (Harvard Business School Press, 2002): the five conditions
• Hackman, interviewed by Diane Coutu, Why Teams Don't Work (opens in a new tab), Harvard Business Review, May 2009: team size, boundaries, deviants
• Ruth Wageman, Debra Nunes, James Burruss and J. Richard Hackman, Senior Leadership Teams (Harvard Business School Press, 2008): research on 120+ top teams
• Gordon Curphy and Robert Hogan, The Rocket Model: Practical Advice for Building High Performing Teams (Hogan Press, 2012): research-based alternative and critique of Lencioni
• Jeff Bezos, 2016 Letter to Shareholders (opens in a new tab): "disagree and commit"