Global employee engagement fell to 20% in 2025, its lowest level since 2020 and the first back-to-back annual decline Gallup has ever recorded. Manager engagement dropped harder, from 27% in 2024 to 22% in 2025, and from 31% as recently as 2022. The conditions most teams work inside are getting worse, not better. That is exactly why the structural side of team performance now matters more than the motivational side.

Most managers who set out to improve their team’s performance start in the wrong place. They look at the people first: who is underperforming, who needs more accountability, who needs a push. Sometimes that is the right read. More often it is not. High-performing teams are usually the product of good conditions, not exceptional individuals, and building those conditions is the manager’s actual job. Gallup’s long-running finding is that managers account for at least 70% of the variance in team engagement, which is another way of saying the environment you create explains most of the result.

This guide walks through the levers that genuinely move team performance. Not the pep-talk levers, the structural and operational ones that let a team do its best work week after week.

Clarity beats effort, every time

The most common team performance problem is ambiguity. People are not sure what they should be working on, how success is defined, or how their work connects to anything that matters. When that fog is present, effort scatters. People stay busy without moving anything important forward.

Google’s Project Aristotle, a two-year study of 180 internal teams, landed on five dynamics that separate effective teams, and two of them are clarity in disguise: dependability (can we count on each other to deliver quality work on time) and structure and clarity (are goals, roles, and plans clear). Before you audit individual performance, audit the clarity your team is operating with:

• Does everyone know the top priorities right now? Not a list of twenty items, the two or three things that matter most this month. If your people give different answers when you ask what matters most, you have a clarity problem, not a motivation problem.
• Does everyone know how success is defined? What does “done” look like, and what does “good” look like? Are those performance expectations specific enough to act on?
• Does everyone understand why? Context is not a nicety. People make better tradeoffs when they know the purpose behind the work, which is the “meaning” and “impact” that Project Aristotle put at the base of team effectiveness.

Clarity does not require a new tool or a new process. It requires a manager who communicates consistently, connects work to outcomes, and keeps the team pointed at what matters. That alone improves performance more reliably than most of the interventions managers reach for first.

Hunt down friction before you add pressure

Friction is anything that makes it harder for your team to do the work: unnecessary meetings, unclear processes, tools that fight back, dependencies that force people to wait on another team, approval chains that slow decisions to a crawl. Friction is dangerous precisely because it is invisible to managers who no longer do the hands-on work. Your team absorbs it daily, and often stops mentioning it because they have learned that mentioning it changes nothing.

Early in my operations career at a large telecom, I ran a weekly review where the same two engineers always looked slammed while the rest of the team had slack. I assumed it was a workload-balancing problem and kept reassigning tickets. It made no difference. When I finally sat with them for an afternoon, the real issue was a single approval step: every change they owned needed a sign-off from a group that answered tickets twice a day, so their work sat idle for hours at a time. The fix was not more people or more motivation. It was moving one approval closer to the team. Their throughput jumped the next week. I had spent a month treating a process problem as a performance problem.

The fastest way to surface friction is to ask, in your one-on-ones: “What slowed you down this week?” “What do you spend time on that does not feel worth it?” “If you could change one thing about how we work, what would it be?” Then act on the answers. Teams that report a bottleneck and watch nothing change learn to stop reporting. Teams that watch their manager actually clear a blocker build the habit of raising the next one. An hour of blocked work removed is worth more than two hours of extra effort applied, so if you suspect time is leaking into rework or messy handoffs, that is where to start.

Psychological safety is the multiplier

Project Aristotle’s headline result was that psychological safety was the single most important of the five dynamics, the foundation the other four stand on. Psychological safety is the belief that you can take an interpersonal risk, sharing a half-formed idea, admitting a mistake, disagreeing with the manager, without being punished for it.

Teams without it are slow and brittle. People do not raise problems early because they are afraid of looking incompetent. They do not challenge a bad call because they are afraid of the conflict. They do not experiment because failure feels dangerous. If your meetings go quiet the moment you ask for input, treat that as a signal, not as agreement.

Managers build psychological safety through consistent behavior over time:

• React to mistakes by understanding them, not by punishing them.
• Actively pull in dissenting views, especially from quieter people, which is where active listening does the real work.
• Model it yourself: say when you do not know something, when you were wrong, when you are still figuring it out.
• Thank people for surfacing problems, even the inconvenient ones.

None of this is soft. It is the structural condition that lets a team perform above the sum of its parts.

Structure decides how fast a team can move

How a team is built shapes how fast it can move and how cleanly it can own outcomes. A team that depends on other teams for every significant decision or handoff will always be slower than a team with the skills and authority to carry work end to end.

If your team constantly waits on others, cannot ship without lengthy coordination, or lacks a key skill that forces you to borrow help every time, that is a structural problem, not a performance problem. The component versus feature team decision determines how much autonomy your team actually has.

Inside your own team, structure means clarity about roles and ownership. Who decides what? Who is accountable for which outcomes? Where do handoffs happen, and how? Ambiguous ownership is one of the most reliable predictors of dropped balls and duplicated effort. Span of control matters here too: stretch one manager across too many direct reports and something gives, usually the coaching and the one-on-ones that hold performance together, and the manager quietly becomes the bottleneck that slows everyone down.

Protect attention, not just hours

Team capacity is not the number of hours on the calendar. It is the cognitive energy available to spend on hard problems. A team that context-switches all day, sits in too many meetings, and fields constant interruptions has far less real capacity than its schedule suggests.

Protecting focus time is one of the highest-leverage moves a manager can make. Cluster the collaborative work (meetings, reviews, syncs) into defined windows. Set norms that defend deep work. Cut the number of things in flight at once, because finishing creates momentum, and momentum drives output more than raw effort does. Watch, too, for the slow drain that comes from chronic stress and work that never feels finished. Teams that perform consistently are not working harder than everyone else. They are working with more focus, more clarity, and less wasted motion.

Hybrid and remote teams still come down to management

Where people work has settled into a stable pattern. Among remote-capable US employees, Gallup puts 52% on a hybrid schedule, 26% fully remote, and 22% fully on-site, with only minor movement over the past two years. So for most managers, leading across locations is not a temporary condition to wait out. It is the job.

The evidence says location is not the performance risk that many leaders fear. The largest randomized trial on the question, run by Stanford’s Nicholas Bloom with 1,612 employees at Trip.com and published in Nature in 2024, found that a two-day-a-week hybrid schedule cut resignations by 33% with no drop in productivity or promotion rates, and employees valued the arrangement as much as a 10% raise. The result did not come from the location itself. It came from managing the arrangement well.

For hybrid and remote teams the fundamentals do not change (clarity, low friction, psychological safety), but execution takes more deliberate effort:

• Over-communicate priorities and context. What used to happen in hallway conversations now has to happen on purpose, in writing, in your async communication.
• Make meetings earn their slot. When face time is scarce, every meeting should end in decisions, not just discussion.
• Build connection on the calendar. Skip-level meetings and steady one-on-ones are not optional overhead; they are how you keep the trust that makes distributed work function.

Where AI actually helps a team perform

The newest variable in team performance is AI, and the 2026 Gallup data shows the manager sitting right in the middle of whether it helps. As of May 2026, Gallup reports 15% of US employees use AI daily in their role and roughly three in ten use it weekly. The tool is spreading fast. The results are not automatic.

The differentiator is management. Gallup found employees are 7.4 times more likely to strongly agree that AI gives them more opportunity to do what they do best when their manager actively supports its use, yet fewer than one in three employees say their manager does. That gap is the opportunity. A manager who treats AI as a capacity lever, naming where it is useful, setting guardrails on where it is not, and clearing the drudge work off people’s plates, turns it into recovered focus time. A manager who ignores it leaves the team to sort out an inconsistent, quietly anxious version on its own. AI does not fix a poorly run team; it amplifies whatever conditions already exist.

Measure a few things that tell you the truth

You cannot improve team performance without knowing what is actually happening, but most managers either drown in dashboards or run on gut feel until something breaks. Track a small set of KPIs that tell you whether the team is heading the right way:

• Output quality and pace. Are you shipping work that meets the bar at a sustainable rhythm?
• Cycle time. How long does work take to move from start to done? A rising cycle time is often the first visible symptom of hidden friction.
• Engagement signals. Are people raising problems, volunteering for stretch work, contributing in meetings? Silence is rarely contentment.
• Retention and stability. Turnover is a lagging indicator, but stay interviews give you a leading one before you lose someone.

The point is not more reporting. It is catching problems early and knowing whether your changes are working. Run after-action reviews on finished projects to learn what actually drove the result versus what you assumed did.

Build the habit of getting better

High-performing teams treat improvement as a deliberate practice, not a happy accident. In most Agile environments the mechanism is the retrospective, a structured look back after each sprint or milestone, but the format matters far less than the habit of carving out time to ask what is working, what is not, and what changes next.

The catch is that a retrospective has to produce action. A team that surfaces the same problems every cycle and changes nothing learns that the exercise is theater and checks out. When something comes out of a retro, follow through visibly, so people see that their input changes how the team operates.

Individual growth feeds the same loop. People who are learning, taking on more ownership, and getting better at their craft raise the team’s ceiling; people who are stagnating either leave or coast. That does not require a formal program. It requires regular one-on-ones where growth is actually on the agenda, stretch assignments that push people just past their comfort zone, and feedback specific enough to act on, tied into how you set expectations and build accountability across the team.

Start with one lever

Improving team performance is mostly about improving conditions. Clarity, less friction, psychological safety, the right structure, protected attention, sane hybrid practices, sensible use of AI, and steady development: those are the levers, and none of them depend on exceptional individuals. They depend on a manager who watches the right things and follows through.

With engagement at a historic low and the load on managers still climbing, the structural work matters more than it ever has. The teams that outperform are not the ones with the most talented people. They are the ones where talented people can actually do their best work, because someone built an environment that makes it possible.

Pick one lever. If you are not sure which, ask your team what slowed them down this week, and then go fix it. That single act of listening and responding will tell you more about your team’s real performance barriers than any dashboard ever will.