The first proposal comes back covered in edits. The second gets a sigh. By the third, the employee has stopped asking what a good version would look like and started asking what will get through the meeting without another lecture.
None of this requires a cruel boss. A leader can want excellent work, care about the team, and still teach people that their effort changes nothing. Authority gives ordinary habits an unusual weight: a private disappointment becomes a team’s standard; an interruption becomes a signal about whose judgment counts.
Leaders make bad calls. They lose patience, miss context, and sometimes talk too much. The important question is what happens afterward. Can the person affected recover, improve, disagree, or make the next decision? These are five lines worth watching because crossing them can make the answer feel like no.
They Don’t Make Approval Unattainable
High standards are useful when people can locate them. “The forecast needs a demand assumption for each region” gives someone a place to work. “This isn’t strategic enough,” repeated after every revision without an example or decision criterion, gives them a mood to manage.
The employee may keep polishing, but the work has quietly changed. Instead of trying to solve a customer problem, she tries to predict the boss’s reaction. Sometimes the boss is right that the work is weak. That makes clear feedback more necessary, not less. Researchers Avraham Kluger and Angelo DeNisi, reviewing hundreds of studies of feedback interventions, found that feedback did not reliably improve performance in every case; its effect depends in part on where it directs attention (their meta-analysis). That finding does not diagnose this particular boss. It is a good reason to be more precise about what someone should change.
If you lead, try naming the standard before the next revision: what decision must this work support, what would be sufficient, and which two things matter most? If you report to someone whose approval moves, bring two concrete alternatives and ask, “Which would you approve, and what makes it stronger?” A specific answer gives you something to build on. An evasive one tells you the obstacle may be the review process rather than your skill.
They Don’t Write People Off
Someone mishandles an account. The leader removes the account, reasonably enough, and assigns a smaller one while the employee rebuilds trust. Six months later, the employee has met the new targets, but the old mistake still decides every assignment. Nothing is said outright; the verdict simply keeps appearing in the work they are allowed to do.
Performance decisions have consequences. Some errors require a long period of supervision, and some roles are a poor fit. Writing a person off is a different act: the leader stops identifying what evidence would change the judgment. Then “prove yourself” becomes an impossible instruction.
The practical alternative is a path back that can be seen from both sides. Name the risk, the work the person can own now, the evidence that would justify more responsibility, and a date to review it. An employee can ask for those conditions directly. If there is no credible path, at least both parties can make an honest decision about the role instead of living under an unstated sentence.
They Don’t Hold Grudges
A purchasing manager misses an order window. The team loses time, perhaps a customer. The manager should account for what happened, repair what can be repaired, and change the ordering process. A leader who keeps bringing up the miss in unrelated conversations has moved from correction into collection.
The distinction is not forgiveness without consequences. A serious failure may permanently change the scope of someone’s authority. But consequences should have a reason and, when possible, a review condition. A grudge has neither. It makes yesterday’s mistake available as an explanation for whatever goes wrong tomorrow.
If you are the leader, say what is still unresolved and what will close it. If you are the person under that shadow, ask for a conversation about the current risk and the conditions for rebuilding trust. You may not control the answer, but you can make the ambiguity visible.
They Don’t Hijack Meetings
The operations lead has barely finished describing a late shipment when the senior leader announces the cause and the fix. Two people know the cause is more complicated. Neither interrupts. The meeting ends with an action plan, and the useful information stays in the room.
Leaders are allowed to have opinions, including strong ones. But an early, confident opinion from the person who evaluates everyone can turn a discussion into a test of agreement. In a study of employee voice, management researchers James Detert and Ethan Burris found that managerial openness was associated with employees speaking up (the study). The research supports paying attention to how the invitation to speak is experienced, rather than assuming that an open calendar or a request for input is enough.
One small change is to ask for the operating picture before offering a theory: “What are we missing?” Then wait, invite the people nearest the work, and distinguish facts from proposals. If you are in the meeting and the leader has already taken over, a useful interruption can be modest: “Before we settle on that, can I show you what happened on the last two orders?” It may not work every time. It puts evidence, rather than status, back into the conversation.
They Don’t Live in a Silo
Engineering has a good relationship with the boss. They speak the same language, have solved hard problems together, and sometimes finish each other’s sentences. Then a customer issue crosses departments. Marketing sees one problem, Sales sees another, and Shipping has the dates that could settle part of the dispute. By the time they meet, the leader has already adopted Engineering’s account of what happened.
Engineering may be right. The problem is that everyone else can see which team’s explanation gets a hearing before the facts arrive. In that moment, a leader’s closeness to one group becomes an organizational boundary. Colleagues learn that they must first overcome an alliance before their information can count. What looks to the leader like trust in capable people can feel to the other teams like the outcome was decided in a conversation they never attended.
The leader does not need to pretend every account is equally well supported. They do need to hear the departments separately, name the decision criteria, and check the handoffs before accepting a convenient story. “What did each team know, and when?” is more useful than “Whose side are you on?” If you are outside the favored group, bring the chronology, the customer effect, and the decision you need. Ask for a shared review of the handoff rather than trying to win a contest of relationships.
Harvard researcher Amy Edmondson’s study of 51 manufacturing teams connected psychological safety with learning behavior (the original study). That study does not tell us which department is right in any dispute. It does offer a useful test of the leader’s behavior: can people raise an inconvenient fact or admit a mistake across team boundaries without assuming that their standing depends on which group they belong to?
These lines are easy to cross in small ways. A tired correction becomes an impossible standard. A justified consequence outlives its reason. An informed opinion arrives before the information, or a trusted team’s version becomes the whole story. The work of leadership is to notice when authority has made those moments larger than intended, then give people a real way to affect what happens next.
People can work with a high standard. They struggle when the outcome has already been decided.