Why Cross-Functional Teams Fail, and the Three Conditions That Fix It
Why Cross-Functional Teams Fail, and the Three Conditions That Fix It
Quick answer: Why do cross-functional teams fail? Leaders build their plans inside their own functions and work across functions without structure. Behnam Tabrizi's study of 95 teams found nearly 75% were dysfunctional, and projects with strong governance support had a 76% success rate. Three conditions help: context sent ahead, agreed rules for disagreement, and small, time-boxed breakouts.
Key takeaways
Nearly 75% of 95 cross-functional teams in 25 leading corporations were dysfunctional (Tabrizi, Harvard Business Review, 2015).
In the same study, projects with strong governance support had a 76% success rate.
A one-page, six-part brief sent 48 hours ahead lets leaders arrive ready to solve the problem.
Authentic dissent produces broader thinking than an assigned devil's advocate (Nemeth, Brown and Rogers, 2001).
Working alone first, then pooling ideas, produced nearly twice as many ideas as group brainstorming (Diehl and Stroebe, 1987).
Plan three to five practice rounds, with an After Action Review after each one.
Watch Mickey Anderson on UNLEASHED: Why Most Cross-Functional Teams Fall Short.
End every practice round with a debrief that changes something
Leaders get better at working across functions when they review each session honestly before the next one. The free After Action Review Agenda Template gives your leadership team a blameless review built on four questions, ending in one real change, so every cross-functional session improves the one after it.
Get the After Action Review Agenda Template
Why do cross-functional teams fail?
Cross-functional teams fail when each function builds its plan on its own, and leaders find out about each other's plans too late to change anything. Every leader can do excellent work for their function while the plans work against each other.
Here's a scene we see in growing organizations.
Marketing plans around lead volume, sales around close rate, and customer success around retention, and the three leaders didn't build those plans together, even though each plan depends on the other two.
Marketing brings in more leads, including some that don't fit.
Sales closes deals with promises that delivery can't keep.
Customer success ends up managing expectations it had no part in setting.
Mickey Anderson, Co-Founder of LoyaltyOps, said on UNLEASHED:
"Each of these functions can hit their targets, but that doesn't necessarily mean the company is going to hit its growth plan."
Each of these leaders is doing good work, and what's missing is structured time to plan and solve problems together, which is a leadership gap your organization can close.
Behnam Tabrizi, writing in Harvard Business Review in June 2015, studied 95 teams in 25 leading corporations and found that nearly 75% of cross-functional teams were dysfunctional. A team counted as dysfunctional when it failed at least three of five criteria:
Meeting a planned budget
Staying on schedule
Adhering to specifications
Meeting customer expectations
Maintaining alignment with the company's corporate goals
The same study found that projects with strong governance support had a 76% success rate. Strong governance meant a senior cross-functional group or a single senior executive champion behind the work, which shows that leadership around the work changes the result.
Nearly 75% of cross-functional teams were dysfunctional, while 76% with governance support succeeded (Tabrizi, HBR).
What changes when leaders build plans together?
When leaders from different functions build a plan together, they see more of the problem, test assumptions with the people who hold the information, and design handoffs before anything breaks.
Better ideas, because each function brings a different view of the customer and the work.
Better plans, because sales can tell marketing which leads close, and customer success can tell sales which promises aren't being kept.
Better solutions, because teams design handoffs together before anything breaks.
Better execution, because the leaders delivering the plan understand the intent behind it and own it.
This applies to annual and quarterly plans as much as to problems. Working across functions is also a separate skill, because leaders are usually promoted for leading their own function well, and few have been taught to share context, disagree well and keep to a time limit with peers from other functions.
Mickey said on UNLEASHED:
"Working across functions is a skill a leadership team builds with practice."
Our post on executive team alignment and the first team covers why each executive's peer group matters as much as their function.
What has to be in place before leaders plan together?
Three conditions let structured time across functions produce a better plan than each function would produce alone: context shared in advance, agreed ground rules for disagreement, and time boxes with small breakouts. Each one is within the leader's control.
Condition 1: Send a one-page brief 48 hours ahead
Leaders can't think well about a problem they're hearing for the first time, and when the context arrives in the session, most of the time goes to briefing and decisions get postponed. A short brief sent 48 hours ahead gives leaders time to reflect, and it often brings out the quieter voices.
A good brief covers six things:
The objective or problem, and why it matters now
What's been tried, and what worked or didn't
What each function is doing about it today, including work that overlaps
Resources available: budget, people, time and tools
Resources not available, stated plainly as constraints
The assumptions everyone is working with that haven't been tested
Make reading the brief a requirement for attending.
A six-part brief for cross-functional planning: objective, past attempts, current work, resources, constraints, assumptions.
Condition 2: Agree the ground rules for disagreement
Constructive objection is where a better plan comes from, as long as it's respectful. Mickey said on UNLEASHED:
"Disagreement is where better ideas come from."
Agree on the ground rules at the start of each session and reinforce them as soon as they slip. Here are some examples:
Challenge the plan and leave the person out of it.
Bring a reason with every objection, and an alternative where you can.
Anyone can name an assumption and ask whether it's been tested.
The most senior leader speaks last.
One person owns the decision at the end, after everyone has had their say.
The Decision Rights Map helps with the last rule, giving each decision one owner and one of four decision types.
Charlan Nemeth and her colleagues, in experiments published in the European Journal of Social Psychology in 2001, compared an assigned devil's advocate with authentic dissent. With an assigned devil's advocate, people mostly generated thoughts that supported their original position, which the researchers called cognitive bolstering. With authentic dissent, people thought more broadly and came up with ideas on both sides of the issue.
Assigned devil's advocates bolstered original views, while authentic dissent produced ideas on both sides (Nemeth, 2001).
An assigned devil's advocate is a reasonable way to start with a team that isn't used to disagreeing, and the ground rules help leaders move toward saying what they actually think.
Amy Edmondson's 1999 study of 51 work teams in a manufacturing company, published in Administrative Science Quarterly, found that teams with psychological safety, a shared belief that the team is safe for interpersonal risk-taking, showed more learning behaviour, which was linked to team performance.
Mickey said on UNLEASHED:
"People raise objections when they believe it's safe to."
A premortem makes objecting part of the task. Gary Klein, writing in Harvard Business Review in September 2007, describes asking a team to imagine the plan has already failed and write down why, and cites 1989 research by Deborah Mitchell, Jay Russo and Nancy Pennington finding that this increases the ability to correctly identify reasons for future outcomes by 30%.
Condition 3: Use time boxes and small breakouts
In one open discussion, only one person can talk at a time, so others wait, lose their ideas or keep them to themselves. Michael Diehl and Wolfgang Stroebe, in the Journal of Personality and Social Psychology in 1987, found that people who worked alone and then pooled their ideas produced nearly twice as many different ideas as the same number of people brainstorming together. Most of the loss came from production blocking, where people wait their turn to speak.
Design each session around that finding:
Give everyone three to five minutes of silent writing on each question.
Move into groups of three or four, mixed across functions, with 15 to 20 minutes on one question.
Have each group report back, and let the whole team decide.
Name a facilitator to keep time and a scribe to capture decisions, owners and next steps.
The Meeting Standards Checklist sets out both roles, with a before, during and after checklist.
How do leaders build cross-functional collaboration with practice?
Leaders build cross-functional collaboration the way they build any skill, with practice on problems that get more real over time and a debrief after every round. Plan three to five practice rounds before the group takes on an objective with real risk attached.
Round 1 is a practice problem. Use a simulated problem or a past one with a known outcome; keep it to 45 minutes, and focus on how the team works together.
Rounds 2 and 3 are small, low-risk problems. Choose something that crosses functions without much weight, so the team can focus on working well together.
Rounds 4 and 5 are real objectives. Work through a planning or strategy objective that matters to the business.
The first session usually feels awkward, and that's a normal part of learning. Debrief every round with the After Action Review, which the US Army introduced in the mid-1970s to capture lessons from simulated battles, as David Garvin describes in Learning in Action (2000). It uses four questions:
What did we set out to do?
What actually happened?
Why did it happen?
What are we going to do next time?
Mickey's advice on UNLEASHED is to "commit to three to five rounds and debrief after each one."
What gets in the way when the sessions start?
The problems that show up in early cross-functional sessions are predictable, and each one has a fix you can plan before the first session.
Leaders arrive without context, and the first 20 minutes go to explaining the problem. Send the six-part brief 48 hours ahead.
The session drifts into status updates. Open with the one question the group is there to answer.
One voice sets the direction early. Use silent writing and breakouts, and have the senior leader speak last.
Disagreement turns personal or goes quiet. Agree the ground rules in advance and use a premortem.
The agenda tries to cover five problems, and none get finished. Assign one problem or objective to each session.
The group leaves without owners. Name one owner and a date for each next step before anyone leaves.
The team judges the approach after one messy round. Commit to three to five rounds with a debrief after each.
How do you start in an organization of 30 to 300 people?
In an organization of 30 to 300 people, start with one group of functions whose work overlaps, one objective, and a standing session every two weeks. Marketing, sales and customer success are a strong first group because they share a customer and rarely share a plan, and finance and operations are another good place to start.
Choose the group where friction between functions costs your organization the most today.
List the problems and objectives those functions already share, with the outcome each one should produce.
Book a 60-minute session every two weeks, and keep it even when one session feels light.
Hold three to five practice rounds with the brief, the ground rules, silent writing and breakouts.
Rotate the facilitator and scribe roles, so the skill belongs to the whole group.
Move the group onto the real objectives on its list.
You'll know it's working when leaders arrive with questions about the brief, objections are aimed at the plan, and handoffs between functions work without a senior leader stepping in. Our page on cross-functional collaboration covers why work stalls at the handoffs. To see how each leader on the team leads today, the Leadership Competency Assessment measures it against clear leadership competencies, with input from the people who see their work.
Frequently asked questions
What percentage of cross-functional teams are dysfunctional?
Behnam Tabrizi's study, published in Harvard Business Review in 2015, looked at 95 teams in 25 leading corporations and found nearly 75% of cross-functional teams were dysfunctional. A team counted as dysfunctional when it failed at least three of five criteria: budget, schedule, specifications, customer expectations, and alignment with corporate goals. Projects with strong governance support had a 76% success rate.
What should a cross-functional meeting brief include?
A good brief covers six things: the objective or problem and why it matters now, what's been tried, what each function is doing about it today, the resources available, the resources not available, and the assumptions that haven't been tested. Keep it to one page, send it 48 hours ahead, and make reading it a requirement for attending.
What is a premortem?
A premortem asks a team to imagine its plan has already failed and write down why. Gary Klein described it in Harvard Business Review in 2007, citing research that imagining an event has already happened increases the ability to correctly identify reasons for future outcomes by 30%. It gives everyone permission to object, because objecting is the task.
How many practice sessions does a cross-functional group need?
Plan three to five practice rounds before the group takes on an objective with real risk. Start with a simulated or past problem, move to small low-risk problems that cross functions, then work on real planning or strategy objectives. Debrief after every round with an After Action Review, and expect the first session to feel awkward.
Help your leaders plan and solve problems across functions
If your functions are hitting their own numbers while their plans work against each other, start with a conversation about how your leaders plan and solve problems together. LoyaltyOps is a leadership development firm that develops individual leaders and leadership teams against clear leadership competencies, so performance improves across your organization.
Sources
Behnam Tabrizi, "75% of Cross-Functional Teams Are Dysfunctional," Harvard Business Review, June 23, 2015. https://hbr.org/2015/06/75-of-cross-functional-teams-are-dysfunctional
Will Yakowicz, "The Simple Reason Your Team Members Don't Work Well Together," Inc., June 24, 2015. https://www.inc.com/will-yakowicz/simple-reason-75-percent-cross-functional-team-projects-fail.html
Charlan Nemeth, Keith Brown and John Rogers, "Devil's advocate versus authentic dissent: stimulating quantity and quality," European Journal of Social Psychology, 2001. https://doi.org/10.1002/ejsp.58
Amy Edmondson, "Psychological Safety and Learning Behavior in Work Teams," Administrative Science Quarterly, June 1999. https://doi.org/10.2307/2666999
Gary Klein, "Performing a Project Premortem," Harvard Business Review, September 2007. https://hbr.org/2007/09/performing-a-project-premortem
Michael Diehl and Wolfgang Stroebe, "Productivity Loss in Brainstorming Groups: Toward the Solution of a Riddle," Journal of Personality and Social Psychology, 1987. https://doi.org/10.1037/0022-3514.53.3.497
David Garvin, Learning in Action: A Guide to Putting the Learning Organization to Work, Harvard Business School Press, 2000.
UNLEASHED, "Why Most Cross-Functional Teams Fall Short," Mickey Anderson, October 1, 2026. https://youtu.be/e9dG2rr2VU4