By Johnny Campbell
Introduction
Empathy at work is not warmth. It’s accuracy. It’s whether someone can correctly predict how a decision will land on people who aren’t in the room, and adjust before it does damage. A manager who announces a restructure on a Friday afternoon because they didn’t think through what a weekend of silence does to a team has low empathy, no matter how kind they were in the meeting. A salesperson who hears “we need to think about it” and correctly reads my finance lead hasn’t been consulted and I’m embarrassed about it has high empathy, even if they never once said anything comforting. Niceness is a style. Empathy is an accuracy rate.
Most interviewers assess this badly, and in three predictable ways. The first is scoring pleasantness in the room rather than perspective-taking on the job — the candidate was warm to the receptionist, made good eye contact, and that becomes evidence. The second is asking about a time they helped someone and accepting an anecdote about being nice: they bought a colleague a coffee, they covered a shift. Kind, but it tests generosity, not perception. The third is never checking whether the candidate’s read was actually right. They say “I could tell she was frustrated” and the interviewer nods, when the only useful follow-up is how did you find out whether you were correct?
What follows is 24 behavioural and situational questions across four capability areas: reading unstated concerns, accuracy of perspective-taking, adjusting under emotional pressure, and empathy without capitulation. Every one is designed to make a candidate produce a specific incident with a specific other person in it, drawn from structured interview research and our own patterns across 10,000+ candidate assessments. Alongside them you’ll find the red flags worth watching for and guidance on scoring answers consistently — which matters more here than almost anywhere else, because empathy is one of the easiest competencies in the world to confuse with liking someone.
Reading Unstated Concerns
What good looks like
A strong candidate notices what hasn’t been said and treats it as real information, not a hunch to be ignored. They can point to a specific signal — a hesitation, a too-quick yes, a question that didn’t get asked — and describe what they did about it before it became a problem: a private conversation, a changed plan, a different sequencing. The evidence is an action taken early, not a feeling reported after the fact.
Behavioural questions
- Tell me about a time you picked up on a concern someone hadn’t said out loud.
- Who was the person, and what was the setting where this happened?
- What specifically did you notice — a word choice, a pause, something that didn’t add up?
- What did you do with it before you knew whether you were right?
- How did you find out afterwards whether you’d read it correctly?
- Walk me through a meeting where the real objection surfaced after it ended.
- Who raised the real objection, and how long after the meeting did it surface?
- What had they said in the room instead, if anything?
- What did it cost you — time, trust, a redone piece of work?
- What would you do differently to surface it inside the meeting next time?
- Describe a situation where someone said yes and you didn’t believe them.
- Who said yes, and what were they agreeing to?
- What told you the yes wasn’t real?
- Did you act on your doubt, or wait to be proven right — what did you actually do?
- What turned out to be true, and how did you find out?
- Tell me about a time you noticed a colleague was struggling before they mentioned it.
- Who was the colleague, and what changed that you picked up on?
- What did you do before they said anything to you directly?
- What did they say when you raised it, and how close was your read to what was actually going on?
- What would have happened if you’d waited for them to bring it to you?
- Walk me through a decision where you anticipated an objection nobody had raised yet.
- What was the decision, and whose objection did you anticipate?
- What told you it was coming before anyone said anything?
- What did you change about the decision or how you presented it?
- Did the objection surface anyway, and how close was it to what you’d expected?
- Describe a time you changed how you delivered something based on who was in the room.
- What was being delivered, and who specifically changed your approach?
- What did you change — the words, the order, what you left out?
- What told you the standard delivery wouldn’t work for that person?
- How do you know the change actually landed better?
Situational scenario
You’re presenting a reorg plan to your team. Everyone nods, no questions, and the meeting ends in twelve minutes instead of the scheduled thirty. Two people who’d normally push back said nothing at all. You have to submit the final structure to your VP by the end of the week.
Walk me through what you do in the twenty-four hours after that meeting.
Then introduce a constraint: “One of the two quiet people reports to a peer of yours, not to you directly, and going to them privately means your peer will know you went around them. What do you do?”
What to listen for
whether they treat the silence as data rather than agreement, whether they find a way to reach the quiet person without triggering the political cost, and whether they still hit the Friday deadline or push back on the timeline itself.
Accuracy of Perspective-Taking
What good looks like
A strong candidate treats their read of another person as a hypothesis, not a fact, and can describe how they checked it. They can name a specific stakeholder whose priorities they worked out correctly under real constraints, and — just as importantly — a time they read someone wrong and found out. Accuracy, not confidence, is the tell: they talk about their hit rate, not their instinct.
Behavioural questions
- Tell me about a time you had to make a decision that would affect people you couldn’t consult.
- Who were the people, and why couldn’t you consult them?
- What did you do instead to work out how it would land?
- What did you get right, and what did you get wrong?
- How did you find out which was which?
- Walk me through how you worked out what mattered most to a stakeholder you barely knew.
- Who was the stakeholder, and what did you actually know about them going in?
- What did you do to find out what mattered to them beyond what they said in the room?
- What was your working theory, and how did you test it?
- Where were you wrong, and what corrected you?
- Describe a time you read someone completely wrong and found out later.
- Who was it, and what did you believe about them at the time?
- What was actually going on that you missed?
- How did you find out you’d been wrong?
- What did you do once you knew — did you say anything to them?
- Tell me about a time you disagreed with someone and could still argue their case properly.
- Who was it, and what were you disagreeing about?
- What was their argument, in their own words, not yours?
- What did they say when you put their case back to them?
- Did arguing their case change your own position at all?
- Walk me through a situation where you predicted how a change would land and got it right.
- What was the change, and whose reaction were you predicting?
- What specifically did you predict, before it happened?
- What did you do differently because of that prediction?
- How do you know you were right, rather than just lucky?
- Describe a time you checked whether your read of someone was accurate.
- Who was it, and what was the read you wanted to check?
- How did you actually check it — what did you do or ask?
- What did you find out — were you right or wrong?
- What would you have done differently if the check had gone the other way?
Situational scenario
You’ve inherited a stakeholder relationship from a colleague who left. The stakeholder has a reputation for being difficult, but you’ve never worked with them directly. Your first project together is time-pressured, and their sign-off is the only thing standing between you and shipping.
Walk me through how you form your read of what this person actually needs, before you’ve built any track record with them.
Then introduce a constraint: “Halfway through, a mutual contact tells you the previous colleague completely misjudged this person — the ‘difficult’ reputation was actually a reaction to being briefed badly, not being briefed at all. Does that change what you do?”
What to listen for
whether they were relying on secondhand reputation rather than their own read, whether they built in a way to test their assumption before it mattered, and whether the new information changes their plan or just their opinion of the person.
Adjusting Under Emotional Pressure
What good looks like
A strong candidate keeps reading the other person accurately even when that person is angry, upset, or directing frustration at them. They can describe a specific moment — being criticised, breaking bad news, sensing a mood shift mid-conversation — and what they changed in real time because of what they picked up. The evidence is a concrete adjustment made under pressure, not a claim of having stayed calm.
Behavioural questions
- Tell me about a time someone reacted far more strongly than you expected.
- Who was it, and what set off the reaction?
- What had you expected instead, and why?
- What did you do in the moment, once you realised you’d misjudged it?
- What did you find out afterwards about what was really driving it?
- Walk me through a conversation where you were being criticised and had to keep listening.
- Who was criticising you, and what was the substance of it?
- What was your instinct in the first thirty seconds, and what did you actually do instead?
- What did you pick up about their state that you wouldn’t have if you’d been defending yourself?
- What changed in the conversation once you kept listening rather than responding?
- Describe a situation where you had to stay useful to someone who was visibly upset.
- Who was upset, and what had caused it?
- What did “staying useful” actually look like — what did you do or say?
- What did you deliberately not do or say, because of the state they were in?
- What was the outcome once they’d calmed down?
- Tell me about a time you were the person who had to break bad news.
- Who received the news, and what was it?
- How did you decide the way to deliver it — timing, wording, setting?
- What reaction did you get that you hadn’t planned for?
- What did you do in response to that reaction, right there?
- Walk me through a moment when you realised mid-conversation that you’d misjudged the mood.
- What told you, mid-conversation, that you’d got it wrong?
- What had you assumed the mood was, and what was it actually?
- What did you change about your approach for the rest of that conversation?
- How did the other person react to the shift?
- Describe a time you chose to say nothing when you had something to say.
- What did you have to say, and who would you have said it to?
- What told you this wasn’t the moment?
- What did you do instead, if anything?
- Did you ever say it later, and what had changed by then?
Situational scenario
You’re mid-presentation to a group of senior stakeholders when one of them cuts in and criticises your approach sharply, in front of everyone. It’s personal in tone, not just about the work. You have fifteen minutes left on the agenda and two more topics to get through.
Walk me through what you do in the next thirty seconds, and then for the rest of the meeting.
Then introduce a constraint: “After the meeting, you learn that person had just come out of a conversation where their own numbers were criticised by their boss. Does that change how you follow up with them?”
What to listen for
whether they can describe what they noticed about the person in the moment rather than just how they felt, whether they protected the rest of the agenda without shutting the person down, and whether the follow-up they choose reflects the fuller picture rather than the version they had in the room.
Empathy Without Capitulation
What good looks like
A strong candidate reads someone accurately and still holds their position when the read tells them to. They can describe a specific instance where understanding someone completely didn’t lead to agreeing with them — a harder message delivered, a decision upheld, feedback given that wasn’t welcome. The distinction they draw, without prompting, is between understanding a person and conceding to them.
Behavioural questions
- Tell me about a time being empathetic meant delivering a harder message, not a softer one.
- Who received the message, and what made the harder version the right one?
- What did your read of them tell you that pointed towards harder rather than softer?
- What did you actually say, as close to verbatim as you can recall?
- How did they respond, and what did it cost you to deliver it that way?
- Walk me through a situation where you understood someone’s position completely and still said no.
- Whose position was it, and what did they want?
- What specifically did you understand about why they wanted it?
- What did you say no to, and how did you explain the no?
- What was the relationship like with them afterwards?
- Describe a time you held a decision that you knew would upset someone you respected.
- Who was the person, and what was the decision?
- What told you it would upset them before you made it?
- Did you change anything about how or when you delivered it because of that? What, specifically?
- What did it cost you in the relationship, and has it recovered?
- Tell me about a time your read of someone told you to push harder rather than back off.
- Who was it, and what were you pushing on?
- What in your read of them told you backing off would be the wrong move?
- What did pushing harder actually look like in the moment?
- What happened, and were you right to push?
- Walk me through an occasion when being accommodating turned out to be the wrong call.
- Who were you accommodating, and what did you give way on?
- What did you tell yourself at the time that justified giving way?
- What went wrong as a result?
- What did you do differently the next time a similar situation came up?
- Describe a time you gave someone feedback they weren’t ready to hear.
- Who was it, and what was the feedback?
- What told you they weren’t ready for it — and did you give it anyway, or wait?
- What was their reaction, and what did you do with that reaction?
- What was different about them or the relationship afterwards?
Situational scenario
A direct report you rate highly has just delivered a piece of work that falls short, and you know exactly why: they took a shortcut you’d explicitly warned against. They’re visibly proud of hitting the deadline. Your own manager is going to see this work in a client meeting tomorrow morning.
Walk me through the conversation you have with them today.
Then introduce a constraint: “They tell you they took the shortcut because they were confident it wouldn't matter, and they still think they were right. Does that change what you say next?”
What to listen for
whether they separate acknowledging the report's pride from addressing the actual gap, whether they name the shortcut specifically rather than softening it into generalities, and whether the fix they land on protects tomorrow's meeting without letting the judgement call go unaddressed.
Red flags to watch for when assessing Empathy
Every example is about being supportive. They comfort, they listen, they check in — but there is no story where reading someone accurately led them to be firmer, more direct, or less accommodating. That’s a candidate who has learned empathy as a customer-service behaviour, not a perception skill, and they will fold under pressure from anyone with a strong opinion.
They have never been wrong about how someone felt. Which usually means they’ve never checked. A candidate with a genuine track record of perspective-taking can name a time they misread a person badly, and can usually tell you how they found out. A flawless record here is not evidence of skill; it’s evidence of no feedback loop.
The other person has no interior life in their story. Ask what the colleague was worried about and you get “they were just being difficult” or “it was a personality thing.” No motive, no pressure the other person was under, no plausible version of events where the other person was reasonable.
They describe it as instinct with no artefacts. “I just read people well.” Real answers contain something you can inspect: a question they asked, a private follow-up conversation, a change they made to a plan, a thing they deliberately didn’t say.
Empathy stops at people like them. Every example involves their own team, their own function, their own seniority band. Nothing about the person on the receiving end of a decision, the customer who never complained, the contractor nobody briefed.
They conflate reading someone with agreeing with them. Every story where they understood the other person ends with them changing their position. Understanding and conceding are different acts, and a candidate who can’t separate them will negotiate away anything you give them.
How to Use These Questions in a Structured Interview
Pick one opener per capability area, ask the same four to every candidate for the role, and score with a 1/3/5 rubric. A 1 is a philosophy or a niceness anecdote — “I always try to put myself in the other person’s shoes.” A 3 is a real incident where they noticed something and acted on it, but with no evidence the read was correct. A 5 is a specific incident where they can state what they thought the other person was feeling, what they did differently as a result, and how they later found out whether they were right — including the times they weren’t. The probes matter far more than the opener here: almost nobody prepares for “what specifically did you see or hear that told you that?” or “how would that person describe the conversation?” Ask each opener once, cleanly, then go down — two questions probed hard produce more usable evidence than six asked flat.
Calibrate before you interview, not after, and be honest about why. Empathy is one of the most affinity-biased competencies on any scorecard: interviewers routinely record “they read people well” when what actually happened is “I liked them,” and the two feel identical from the inside. Score against the evidence in the answer, not the feeling in the room, and get your panel to score the same recorded answer independently before comparing — if two interviewers are a band apart, one of them is scoring rapport. See how Interview Intelligence solves this → by capturing how different interviewers frame the same behaviour, so you catch those blind spots before they cost you.
Frequently asked questions
What's the difference between empathy and sympathy in a work context?
Sympathy is feeling something on someone's behalf. Empathy is accurately modelling what they're experiencing and what it means for the decision in front of you. Sympathy produces comfort; empathy produces adjustment. The practical test in interview is whether the candidate's read changed anything — the sequencing of an announcement, the framing of a decision, the person they briefed first. If the story ends with them feeling bad for someone and nothing else happening, that's sympathy.
Isn't empathy just agreeableness by another name?
No, and conflating the two is the most common scoring error on this competency. Agreeableness is a preference for harmony; empathy is a perception skill. They come apart most visibly under pressure: an agreeable person concedes to avoid discomfort, while an empathetic person may read the situation accurately and conclude they need to be firmer. The Empathy Without Capitulation questions exist specifically to separate them.
Can empathy actually be developed, or is it fixed?
The perception element improves substantially with practice, feedback and exposure to people unlike yourself. What rarely shifts is the underlying inclination to bother — some people simply don't attend to other people's states, and no amount of training makes them curious. In hiring terms, look for evidence of a feedback loop: candidates who check their reads, get corrected and adjust are the ones who improve. Candidates with a flawless self-reported record generally don't.
How do I assess empathy for a role with no direct reports and little client contact?
Every role has people downstream of its decisions — the team who inherits the code, the colleague who picks up the handover, the user who never files a complaint. Identify that group before you interview and anchor your questions to it. 'Tell me about a time you anticipated an objection nobody had raised yet' works as well for an engineer as it does for a people manager, because both are being asked whether they thought about someone who wasn't in the room.
Should I use a hypothetical, like 'what would you do if a colleague seemed upset?'
No. Hypotheticals test whether a candidate knows the socially approved answer, which for empathy almost everyone does. Behavioural openers force a specific incident with a specific person, and the details are what you score. If a candidate genuinely has no relevant example, that's information — but ask twice, in two different framings, before you conclude it.