How to Turn Around an Underperforming Team

When a team misses its numbers for the third quarter in a row, most managers reach for the fix they already know. A former engineer schedules training. A former salesperson runs a motivation push. A former consultant redraws the process. Each fix is sensible on its own terms, and each one is a guess.

An underperforming team is usually failing at one of four layers: skills, motivation, process, or how the team behaves once pressure arrives. The first three get audited all the time. The fourth almost never does, and it is often where the trouble sits. Turning the team around starts with finding the failing layer, before any fix goes on the calendar.

Why the First Fix Usually Misses

Gallup estimates that managers account for at least 70% of the variance in employee engagement across business units. That 2015 estimate covers engagement scores specifically. It still helps explain why a struggling team so often reads as a manager problem.

The people expected to lead the recovery are running low themselves. Gallup's State of the Global Workplace research found that manager engagement fell from 27% in 2024 to 22% in 2025, the largest one-year drop it has recorded.

The standard lever is individual performance management, and its track record is poor. Only 2% of Fortune 500 CHROs strongly agree that their performance management system inspires employees to improve. Reviews examine people one at a time. A team can be made of capable individuals and still fail as a unit, and a review cycle has no way to see that.

The Four Layers of an Underperforming Team

Work through these in order.

1. Skills

The question: does the team lack a capability the work now requires? A new stack, a new market, or a senior person who left and took the knowledge with them. Look for work that stalls at the same technical step every time, or for one person who has become the only route through a problem.

Skills are often the first layer managers blame, and they often explain less than expected. In a widely cited Science study, Woolley and colleagues found that a group's collective intelligence moved with members' social sensitivity and equal turn-taking in conversation, with only a weak link to the average or maximum intelligence of its members. Later work has challenged that finding and credited individual IQ instead, so treat it as contested. It should still make you slower to conclude that hiring or training alone will fix the problem.

2. Motivation

The question: has the team stopped caring about the outcome? Signs include missed commitments with no visible frustration, silence in planning, and people doing the minimum the ticket describes. Motivation problems are common. Gallup reports that global employee engagement fell to 20% in 2025, its lowest level since 2020.

A motivated team can still underperform, and pressure is one reason. Heidi Gardner's study of 78 audit and consulting teams found that performance pressure raised team motivation while making teams worse at using their own expertise. If your people are working late and still missing targets, look next at process and at how the team handles pressure.

3. Process

The question: does everyone know what is expected, who decides, and how work moves? Gallup found that just 46% of U.S. employees clearly know what is expected of them at work, down 10 points from a high of 56% in March 2020. Gallup has also found that gains on that single item tend to come with productivity gains of 5% to 10%.

Test it directly. Ask each person to write down the team's top three priorities and who owns the next decision on each. If the answers diverge, you have found a process failure, and it is the cheapest layer to fix. Google's Project Aristotle lists structure and clarity among its five dynamics of effective teams.

4. Behavior Under Pressure

This is the layer almost nobody measures. It covers what the team does when a deadline slips or a plan breaks: who speaks, whose expertise gets used, and whether mistakes surface early or stay hidden until the demo.

Gardner's teams show the mechanism. Under pressure they leaned on general, visible expertise and discounted members' specialist knowledge, and performance fell even as effort rose. Amy Edmondson's study of 51 work teams found that psychological safety predicted learning behaviors (asking questions, admitting errors, seeking feedback), and that those behaviors carried the link to team performance. Project Aristotle, which studied more than 180 Google teams, reached a similar conclusion: how the team worked together counted for more than who was on it, and psychological safety led the list.

In software, the DORA research program found that teams with generative cultures had 30% higher organizational performance, where generative means high trust and open information flow.

The trouble is visibility. This behavior shows up in live moments, forty minutes into an incident or the night before a launch, while engagement surveys record how people feel and individual reviews record what each person produced. For a list of symptoms to compare against, start with these signs your all-star team has no chemistry.

Run the Diagnosis Before the Fix

Two weeks of diagnosis beats a quarter spent on the wrong fix. A workable sequence:

1. Pull the output data. Cycle time, missed commitments, rework, pipeline, or whatever the team is judged on. Note when the drop started and what changed around then. The guide to measuring team performance beyond velocity covers which metrics hold up.
2. Run the clarity test. Priorities and decision owners, written independently, compared side by side.
3. Hold short one-on-ones. Ask what slows each person down and what they would change. Gallup found that nearly half (45%) of employees who quit said neither a manager nor another leader had talked with them about their job satisfaction, performance, or future in the three months before they left.
4. Observe the team under load. Sit in on the next incident review, launch meeting, or deal review. Count who talks. Note whose expertise gets overridden and how long bad news takes to surface.
5. Check for one person. Sometimes the failing layer is a single member. Felps, Mitchell, and Byington describe how one person who withholds effort or breaks norms can spread withdrawal and distrust across a group. Rule the team layer in or out first, then decide whether the issue is individual.

For a structured baseline, the free team dysfunction assessment scores trust, conflict, commitment, accountability, and results in 15 questions.

Match the Fix to the Layer

• Skills: pair people, bring in targeted training, or hire. Set a date to check whether output moved.
• Motivation: reconnect the work to outcomes people care about, remove the obstacles they have flagged, and recognize effort in public. Unaddressed motivation problems tend to turn into turnover. Gallup puts the cost of replacing an employee at one-half to two times their annual salary, which it calls a conservative estimate.
• Process: write down priorities, owners, and who makes which calls. Revisit them every sprint or cycle until the clarity test comes back consistent.
• Behavior under pressure: this layer needs practice. Run blameless reviews after misses. Rotate who leads when a deadline tightens. Ask the quietest specialist to speak first in high-stakes meetings. Then give the team repeated, low-stakes chances to work through pressure together, so the new habits exist before the next crunch.

Whatever the layer, pick one fix, set a measure, and hold it steady for several weeks. Stacking four fixes at once makes it impossible to learn which one worked. The research on what makes a high-performing team points in the same direction: effective teams develop through repeated work together.

Following the Turnaround Week to Week

The hardest part of a turnaround is knowing whether the fourth layer is improving. Output metrics lag by months. Surveys arrive quarterly. You need a weekly read on how the team coordinates.

QuestWorks is one option built for that job. Once a week the whole team plays a 25-minute quest at the same time, seated dynamically into groups of 3 to 6. The pressure comes inside a game, so people fall back on their usual habits: who steps up, who adapts, and whose expertise gets used as the clock runs down. Those gameplay patterns feed a Team Intelligence score from 0 to 100, with an 8-week trend, named dimensions such as Role Clarity and Decision Velocity, and a "Do this week" list of concrete actions. For a turnaround, the trend line shows whether your chosen fix is changing how the team works together.

The tradeoffs: it takes 25 minutes of calendar time every week, the trend needs several weeks before it says much, and it will not solve a skills or staffing problem on its own. Participation is voluntary and the data stays out of performance reviews, which keeps people playing as themselves. Plans start at $199 per month per team with 10 seats included, and the 30-day free trial covers four game days with no credit card required.