How to Choose a Response When a Colleague Is Upset

How can you choose a response when your colleague is upset?

When you do not know what help a colleague wants, begin with what you can observe, ask an open question they can decline, and wait for the answer. A brief check-in gives them room to say whether they want to talk, need practical help, or would rather have space. It is an invitation to clarify, not a claim that you know why they are upset or what will make things better.

Keep your first move small. You might say, “You seem quieter since the handoff meeting. Would it help to talk, look at the next step together, or take some time?” Then leave room for an answer, including “I’m fine” or “I don’t want to get into it.” Their response helps you choose what to do next; it does not settle every work question. A deadline, responsibility, or immediate safety concern can require action before a fuller conversation, but ordinary uncertainty is a reason to ask rather than guess. Avoid turning the check-in into a demand for an explanation. Your colleague may be upset about the work, something outside it, or may not want to name a cause. You do not need to identify the cause before offering a respectful choice. If they decline, acknowledge that and let the conversation pause; if the work still needs attention, return to the concrete task separately.

Why can the helper’s goal lead to the wrong first move?

A helper’s first impulse often serves a goal of its own. You may want to ease the colleague’s distress, solve the work problem, stop an uncomfortable silence, or keep a meeting moving. Those aims can be understandable. But they describe what you want the interaction to accomplish; they do not tell you what the colleague wants from you.

Researchers use the term extrinsic emotion regulation for efforts to influence another person’s emotions. At work, that can mean trying to help someone feel calmer or more hopeful, whether through listening, reassurance, or a change to the situation. The term describes the direction of the effort, not whether it is welcome or effective. A well-meant attempt to improve someone’s mood can still miss if they need a concrete task decision, time to think, or simply a listener who will not rush them.

Illustration: a colleague says a project handoff went badly. You immediately start explaining a fix, perhaps because you want to reduce the pressure and get the work back on track. Before offering it, pause long enough to notice your own aim: are you responding to a request, or trying to make the discomfort disappear? That question can keep an unsolicited solution from taking over the conversation. You can then ask what would be useful and hear whether the problem is the handoff itself, the next decision, or something they do not want to discuss now.

A task-focused aim may be exactly what the moment needs when a decision is time-sensitive. If a deadline is close, identify the decision and ask whether your colleague wants help with that immediate step. The point of checking your motive is not to judge it or suppress it; it is to keep your urgency from standing in for their preference. A useful distinction is whether you are trying to change the situation they described or change how they feel about it. Sometimes those can happen together: a clear next step may reduce uncertainty. Sometimes they diverge: reassurance can sound like pressure to feel better, while jumping into task repair can leave the colleague’s concern unheard. You cannot resolve that difference by assuming the kindest intention will land as intended. A short pause helps you choose a question before choosing an intervention. It keeps help responsive to the conversation as it unfolds.

What can you learn before choosing a response?

A check-in should gather two things: what your colleague is willing to share, and what kind of response they want. Their expression or silence is an observable cue; it does not tell you its cause. “You seem quieter after that meeting” names what you noticed. “You’re angry because the handoff went badly” supplies a story they have not confirmed. Keeping those separate leaves the colleague room to correct you, explain something else, or say nothing.

The Open University teaching resource “Effective communication in the workplace: Dealing with an angry customer or colleague” uses an angry-colleague example to teach respectful communication and open questions about what is driving an exchange. It offers a practical model for inquiry, not evidence that a particular question will calm someone. You could say, “Would you like to talk, get practical help with something, or have some time?” Those are options, not a menu they must complete. Add, “It’s fine if you’d rather not talk about it,” and make that permission credible by accepting a no without asking why.

Then listen for the request, not a diagnosis. If the colleague says, “I need ten minutes before we decide,” you have a preference and a next step; you do not need to test whether the delay reflects frustration, embarrassment, or something outside work. NAMI’s “How to Have Mental Health Conversations” recommends listening, allowing time and privacy, and asking before moving into problem-solving in mental-health conversations. That guidance supports asking permission as a reasonable conversational choice; it does not establish effects for ordinary workplace upset. In practice, one open follow-up—“What would be useful right now?”—is enough. Avoid a string of why-questions, labels, or guesses disguised as concern. They can make a check-in feel like an inquiry the colleague has to pass.

An open question cannot guarantee disclosure or a clear answer. Your colleague may not want to explain, may be unsure what would help, or may prefer to keep the interaction focused on work. Treat that response as information too: acknowledge it, leave the door open without pressing, and move to the concrete decision only if one is needed. The distinction matters when the cue is ambiguous: a quiet response might mean the person is concentrating, choosing words, or simply finished speaking. Because the cue underdetermines the cause, a question should leave several answers possible. If they correct your observation—“I’m quiet because I’m reading the brief”—accept the correction and return to the work. You have learned what you needed without asking them to account for an emotion. The distinction matters when the cue is ambiguous: a quiet response might mean the person is concentrating, choosing words, or simply finished speaking. Because the cue underdetermines the cause, a question should leave several answers possible. If they correct your observation—‘I’m quiet because I’m reading the brief’—accept the correction and return to the work. You have learned what you needed without asking them to account for an emotion.

Sources: Effective communication in the workplace: Dealing with an angry customer or colleague; How to Have Mental Health Conversations

Which response fits what the colleague asks for?

Once the colleague answers, let that preference shape your next move. If they want to be heard, give them your attention and reflect the point you understood before offering an interpretation. If they ask for practical help, clarify the task and what part they want you to take. If they want space, agree on whether you should check back and when. A request for perspective is an invitation to offer it; keep it brief enough that they can accept, question, or set it aside. These responses differ in observable ways, even though each begins with the same act: taking the request seriously.

Illustration: a colleague says they want help prioritizing a handoff. You might ask which item is due first, what is already committed, and what you have capacity to handle. Then agree on one next step: for example, you take the status update while they confirm the revised delivery time. State who will do what and when you will reconnect. That makes the offer useful and bounded. It also avoids turning a request for task help into unsolicited coaching about how the colleague should manage pressure or communicate better.

The preference does not erase the work around it. There may be a deadline, a limit on what you can take on, or uncertainty about who owns a decision. Say that plainly while making the next step specific: “I can review the first part now, but I can’t own the full handoff. Let’s confirm who has the remaining update by three.” If the colleague wants to talk but a work decision cannot wait, separate the two needs: make the necessary decision concise, then agree whether there is time to return to the conversation. A clear boundary is easier to work with than an open-ended offer you cannot keep.

Sometimes the answer will be unclear or change as the conversation continues. Check what you heard in a short sentence—“So you want me to listen now, and we’ll sort the task after lunch?”—and let them correct it. If no support is wanted, respect that and handle any shared work obligation directly. The aim is not to produce the perfect response from a single answer; it is to keep the next action understandable to both people. A request also has a scope. “Can you help?” may mean taking one small action, thinking through options together, or covering a responsibility. If you are unsure, ask one bounded clarifying question before committing: “Would it help if I checked the figures, or are you asking me to take over the update?” This keeps the colleague in control of what they disclose while making the work agreement precise. If their answer is “just stay with me while I decide,” do that rather than layering on advice. A request also has a scope. ‘Can you help?’ may mean taking one small action, thinking through options together, or covering a responsibility. If you are unsure, ask one bounded clarifying question before committing: ‘Would it help if I checked the figures, or are you asking me to take over the update?’ This keeps the colleague in control of what they disclose while making the work agreement precise. If their answer is ‘just stay with me while I decide,’ do that rather than layering on advice.

What do workplace studies say about listening and pressure?

The clearest workplace evidence favors receptive listening as a reasonable first response when the right kind of help is uncertain. In “Does it matter why we try? How goal-focused extrinsic emotion regulation influences co-worker's relational outcomes,” two randomized online experiments asked working adults to choose strategies for influencing a coworker’s emotions. Random assignment lets the researchers test whether a stated goal shifts strategy selection. In the first experiment, participants assigned a goal of improving the coworker’s feelings selected receptive listening more often than participants in the comparison goal conditions. The outcome was what people said they would choose in a survey situation; the experiment did not observe a real upset colleague or measure whether the choice helped. The paper reports samples of 216 and 471 working adults across the two experiments. Their randomized conditions address a narrow question about the goal behind a response, rather than whether one strategy is best across all situations.

The same paper then changes method. Its two-source field study paired 143 employee regulators, who reported their goals and strategies, with 277 coworker targets, who reported relationship outcomes. Receptive listening was associated with lower coworker-rated relationship conflict. Because the strategy and outcome came from different coworkers in the pair, the association is more informative than asking one person to rate both their own listening and the relationship. It still describes a relationship between reports gathered in the field, not a tested intervention: a less conflictual relationship could make listening easier, and other features of the relationship could shape both.

The field results also distinguish listening from expressive suppression. Employees’ reports of trying to suppress coworkers’ emotional expression were associated with higher coworker-rated relationship conflict and lower team-member exchange. Team-member exchange captures the perceived quality of a coworker relationship in the team; it is not the same outcome as conflict. The findings therefore do not collapse into a claim that every effort to influence a colleague is harmful. They identify different reported strategies and different relational outcomes, with the suppression associations pointing in an unfavorable direction.

Taken together, the paper supports a practical default under uncertainty: make room for the colleague to explain what is happening before trying to manage how they feel. Listening leaves the response open to correction; urging someone to hide or contain visible emotion can make the relationship itself harder. That is a defensible judgment from strategy-choice experiments and coworker associations, not an instruction to listen at length in every situation. A colleague may ask for a concrete work decision, need quiet, or decline discussion, and the study did not compare those response scripts.

The boundary matters in the moment. Neither the online experiments nor the observational field study tested enacted responses to a colleague who was upset in a live workplace exchange. The experiments show how goals influence selected strategies in survey scenarios; the field study links reported strategy use with coworkers’ later relationship ratings. Neither establishes that listening caused lower conflict, that suppression caused poorer exchange, or that either response will settle a task disagreement. Use the evidence to choose an open first move, then let the colleague’s answer and the work at hand guide what follows.

Sources: Does it matter why we try? How goal-focused extrinsic emotion regulation influences co-worker's relational outcomes

Why do role and perceived sincerity change the evidence?

The evidence changes when the person responding has authority over the person who is upset. “Perceptions of leader interpersonal emotion regulation: Daily and longitudinal consequences for employees” is a separate 2026 leader–employee study, not part of the coworker paper. Its diary analysis followed 193 employees across 934 workday observations. Employees’ perceptions that a leader was trying to improve their affect were associated with greater same-day positive affect and higher perceived leader effectiveness. Those are daily associations in a hierarchical relationship: the response may feel supportive partly because it comes from someone who can affect the employee’s work, while that same authority can make the employee more attentive to the leader’s motives.

Perceived sincerity helped explain why the leader response did not land uniformly. In the diary analysis, the association between the leader’s affect-improving behavior and perceived effectiveness was weaker when employees saw the behavior as inauthentic, becoming nonsignificant at high perceived inauthenticity. The authors also report separate longer-term and vignette evidence; those designs are distinct from the daily observations and should not be merged into one outcome. For a peer, the narrow implication is that an effort to improve someone’s mood may be received differently depending on whether it seems responsive or managerial. The study cannot tell a coworker how to make that judgment disappear, and a peer does not hold the same role as a supervisor. A supervisor’s attempt to cheer someone up can carry an implicit expectation about how that employee should appear at work. Perceived inauthenticity is therefore part of the reported reception of the behavior, not a clinical judgment about the leader’s character.

A different kind of evidence comes from “Interpersonal emotion regulation strategies: Enabling flexibility in high-stress work environments,” a qualitative study of police patrol work. Researchers observed three two-officer patrol units for four days each and conducted semistructured interviews. Their account describes officers shifting strategies as events unfolded, with interaction between partners and interference from third parties shaping what became possible. It offers a close view of how emotion regulation can change with an operational situation. It does not compare response effectiveness or estimate how common any strategy is among ordinary coworkers.

The contrast is useful because the settings make different demands. A leader’s affect-focused response is filtered through authority and an employee’s reading of sincerity; police patrol involves high-stress events, joint action, and people outside the pair who can alter the interaction. Both sources make a fixed “say this when someone is upset” formula look too simple, but for different reasons. The leader study links recipient perceptions to daily reports; the police research traces a process as circumstances change. Neither supplies a tested peer script for an ordinary disagreement or a colleague’s bad day. Their methods also prevent a simple ranking: a diary can connect day-level perceptions and affect over repeated workdays, while observation and interviews can show how a strategy changes in a particular sequence without showing that the change improved the outcome.

For the peer conversation, take these findings as a reason to attend to how the response is received and to remain ready to adjust, not as a borrowed rule. A question such as “Would talking help, or would you rather have space?” gives a colleague room to signal what fits without asking them to endorse your interpretation. If the exchange is shaped by formal authority, urgent operations, or other people’s actions, the evidence from these different settings cautions against assuming that the same emotional gesture carries the same meaning. The studies illuminate context; the colleague’s response still supplies the relevant information for this particular interaction.

Sources: Perceptions of leader interpersonal emotion regulation: Daily and longitudinal consequences for employees; Interpersonal emotion regulation strategies: Enabling flexibility in high-stress work environments

When do work duties or limits take priority?

A colleague’s preference matters, but it does not erase a deadline, a responsibility you own, or the limits of your role. The National Mental Health Commission’s *Mentally healthy workplaces: Managing change and disruption guide* addresses worker needs in organizational disruption. That is useful context for remembering that support is shaped by what the workplace and each person can actually provide; it does not establish which sentence one coworker should use or whether a particular response will help. The practical judgment here follows from the situation: be clear about the constraint, offer what is feasible, and avoid implying that the colleague must hide an emotion to keep work moving.

Suppose a colleague wants to talk through a difficult exchange, but you have ten minutes before a client handoff that you are responsible for. You could say, “I can give this ten minutes now, then I need to send the handoff by three. I can help check the key details first; would you like to talk after the deadline, or set a time tomorrow?” The wording names the task and the time available, makes a concrete offer, and gives the colleague a choice about continuing later. It does not require them to pretend they are fine or turn a work limit into criticism of their reaction.

A capacity limit deserves the same directness. If you cannot take on another task, say what you can do, if anything: “I can review the first page, but I can’t own the full revision today. The project lead will need to assign the remaining work.” If that means the preferred help is unavailable, acknowledge the disappointment and identify a feasible next step or the person whose role includes deciding priorities. Avoid volunteering time, authority, or follow-up you do not have; a reassuring promise that later proves false can leave both the task and the relationship harder to repair.

If the situation raises an immediate safety concern, follow the established workplace procedures and role responsibilities that apply there. A peer conversation is not a substitute for those procedures, and this article cannot specify a protocol for every workplace. For ordinary distress alongside urgent work, the useful boundary is narrower: state what must happen now, what support you can provide within that limit, and whether there is a realistic time to return to the personal conversation. The concern remains heard even when the task has to come first.

Sources: Mentally healthy workplaces: Managing change and disruption guide

How can you adjust if your first response misses?

A correction from the colleague gives you better information about what they want now. If you offered advice and they say, “I don’t need a solution; I just want you to hear me,” stop making the case for your advice. Acknowledge the mismatch plainly: “I moved into fixing it when you wanted me to listen. I’m sorry I did that. Do you want to tell me what feels most important?” Then listen for the answer instead of asking them to reassure you that your intentions were good. This is an illustrative exchange, not a research-tested repair script.

The apology is for an identifiable action—moving to advice after the colleague signaled a different need—not for failing to guess correctly at the start. You can be attentive without claiming that you should have known what they wanted. A proportionate repair names what you did and lets the colleague say whether they want to continue. If they do, give them room to speak; if they ask to stop or take space, respect that request rather than trying to complete the conversation for your own comfort.

If the colleague corrects a factual detail rather than the kind of support they want, make that distinction explicit. “Thanks for correcting the date; I had it wrong” repairs the factual error. “I understand why the delay put pressure on your part of the work” acknowledges impact. Neither sentence requires deciding, before checking, whether every interpretation of the event is accurate. Where the two are tangled, ask one clarifying question and identify what you can verify, instead of treating a disagreement as proof that the colleague’s feelings are mistaken.

Keep any work disagreement separate from the acknowledgment. You might accept that your quick advice interrupted them while still needing to check whether the handoff was actually late or who owns the next step. Acknowledging impact does not settle those facts. You can say, “I can see that jumping in felt dismissive. I still need to verify the timeline before I say what happened; let’s check that after I’ve heard you.” This makes the interpersonal repair specific while leaving the operational question open for evidence.

The adjustment has a clear limit: adapt when the request is feasible, and say honestly when it is not. If the colleague wants listening and you can stay, listen. If you have to return to a deadline, explain the limit and offer a real time or another appropriate route for the task. If you cannot provide the requested help at all, say so without dressing a refusal as agreement. The colleague’s correction should change your response where you can change it; it does not require you to abandon a necessary decision or promise capacity you lack.

Choose one behavior to practice next

After the conversation, choose one action you can recognize next time. Did you ask what would help before offering a fix? If your colleague asked for space, did you leave room without pressing them to explain? If they corrected your first response, did you adjust rather than defend the intention behind it? Pick the moment that felt least natural and name the behavior you want to try: ask first, pause before advising, or acknowledge and change course. A specific action gives you something observable to notice in the next exchange. You might jot down the cue that helped you notice the choice, such as a direct request, a pause, or a correction, then compare it with what you actually did. Keep the note about your own response, not a judgment about whether your colleague handled the moment correctly.

For a more structured look at your own patterns, the private 32-item Emotional Skills Profile offers recent-behavior reflection alongside judgments on authored scenarios. It can help you examine how you tend to respond; it cannot tell you what a particular colleague wants or guarantee how a conversation will go. [Explore the Emotional Skills Profile](/assessment). If you have already completed it, [review your report](/report) as a prompt for reflection. In the next conversation, the colleague’s words and the situation still guide what you do.

Questions readers ask