Situational interview questions
Situational interview questions pose a hypothetical — "what would you do if…" — and test judgment on situations the candidate hasn't faced yet. Use the 13 below where past experience can't exist: career changers, new processes, scenarios unique to you. Everywhere else, pair each hypothetical with a follow-up asking for a real occasion.
All 13 questions
Common situational interview questions
- 01It's Wednesday. Two deliverables are both due Friday and you can only finish one. Walk me through your next hour.
- 02You spot an error in work your team already sent to a client. Nobody else has noticed. What do you do?
- 03Halfway through a project, the requirements change completely. What's your first move?
- 04A new process from another department slows your team's work noticeably. What do you do in the first two weeks?
- 05You discover a teammate promised a customer something your product can't deliver. The customer doesn't know yet. What do you do?
Situational questions about working with AI
- 06A teammate hands you work to review that's clearly AI-generated and contains a factual error a client would catch. How do you handle it?
- 07You find an AI tool that would save you five hours a week, but using it means uploading client data, and company policy doesn't mention it. What do you do?
- 08An AI assistant gives you an answer that contradicts what a senior colleague told you last week. The work is due today. How do you resolve it?
- 09A client asks whether AI was used to produce the deliverable you just sent. It was, partly. What do you say?
Situational questions about people and priorities
- 10A teammate you depend on is visibly struggling and your shared deadline is at risk. They haven't asked for help. What would you do?
- 11In a meeting, a colleague presents your idea as their own and it lands well. What do you do in the moment, and afterwards?
- 12You inherit a task from someone who left the company. Their notes make no sense and it's due in three days. What's your approach?
- 13Your manager hands you a goal you privately believe is impossible in the time given. What do you do in the first day?
Questions 1–5
Common situational interview questions
It's Wednesday. Two deliverables are both due Friday and you can only finish one. Walk me through your next hour.
A strong answer starts with impact — which miss costs more — then a call to the other owner the same day, with a smaller alternative offered. Weak answers start with working the weekend.
What to look for
- Ranks by cost of the miss, not by which task is easier or louder.
- Contacts the losing deliverable's owner immediately, with an alternative in hand.
- Asks whether either deadline is actually movable before sacrificing one.
Example answerRed flags
Example answer
First I'd check which Friday is real — deadlines often have slack nobody's claimed. If both hold, I'd finish the one tied to revenue or a customer commitment and call the other owner Wednesday afternoon with what I can deliver instead: a draft, a summary, or a firm Tuesday date. What I wouldn't do is go quiet and hope.
Red flags
- Plans to hit both by working nights — the scenario said that isn't an option.
- Chooses by which stakeholder is scarier rather than which miss costs more.
You spot an error in work your team already sent to a client. Nobody else has noticed. What do you do?
Strong candidates verify the error, tell the client before the client finds it, and bring the correction with the confession. The order matters: check first, disclose fast, fix visibly.
What to look for
- Verifies it's actually an error before raising alarms.
- Discloses to the client proactively, not after they notice.
- Pairs the admission with the corrected version, not just an apology.
Example answerRed flags
Example answer
I'd confirm the mistake with whoever built that section — quietly, no blame yet. If it's real, I'd tell my manager and the client the same day: here's what's wrong, here's the corrected file, here's why it happened. Clients forgive a fixed error in hours; they don't forgive discovering it themselves a month later.
Red flags
- Would wait to see if the client notices.
- First instinct is finding who to blame rather than fixing the deliverable.
Halfway through a project, the requirements change completely. What's your first move?
The best answers pause before rebuilding: confirm what actually changed and why, salvage what still applies, and re-estimate openly. Diving straight into rework without re-scoping is the trap.
What to look for
- Asks why the requirements changed — the reason shapes the response.
- Inventories what survives before declaring everything lost.
- Communicates the new timeline cost instead of silently absorbing it.
Example answerRed flags
Example answer
First, a conversation, not a rebuild: what changed, why, and is it final. Then I'd sort the finished work into keep, adapt, discard — usually more survives than the first panic suggests. Then the new estimate, with the change as the stated reason. Absorbing a re-scope silently just teaches people that changes are free.
Red flags
- Starts rebuilding immediately without asking why the change happened.
- Treats the change as a personal affront and relitigates the old requirements.
A new process from another department slows your team's work noticeably. What do you do in the first two weeks?
Good answers comply first, measure the actual slowdown, then bring numbers to the process owner with a proposed adjustment — challenging it with data rather than ignoring it or just venting.
What to look for
- Follows the process while questioning it — no unilateral workarounds.
- Quantifies the cost before complaining: hours lost, tickets delayed.
- Takes the case to the owner with a specific fix, not general frustration.
Example answerRed flags
Example answer
Weeks one and two, we follow it and I log the cost — say, every request now waits two extra days. Then I'd book time with the team that owns the process, show the numbers, and propose an exemption or a lighter path for routine cases. Maybe there's a reason for it I can't see. Maybe not.
Red flags
- Would quietly work around the process from day one.
- Escalates emotionally with no measurement of the actual impact.
You discover a teammate promised a customer something your product can't deliver. The customer doesn't know yet. What do you do?
Strong answers separate the two problems: the customer needs a fast, honest correction with options; the teammate needs a direct private conversation. Letting the promise ride to avoid awkwardness fails both.
What to look for
- Talks to the teammate first to confirm what was actually promised.
- Corrects the customer quickly, with alternatives rather than a bare no.
- Treats the teammate as someone to fix it with, not report on.
Example answerRed flags
Example answer
I'd start with the teammate: show me the thread, maybe I'm misreading it. If it's a real overpromise, we go to the customer together within a day — "we misspoke, here's what we can do instead" — ideally before their plans depend on it. Separately, I'd want to know whether they didn't know or hoped nobody would check.
Red flags
- Would let the customer find out at delivery time.
- Goes straight to management without ever speaking to the teammate.
Questions 6–9
Situational questions about working with AI
2026 · AIA teammate hands you work to review that's clearly AI-generated and contains a factual error a client would catch. How do you handle it?
The strong answer fixes the deliverable first, then addresses the workflow with the teammate directly: AI drafting is fine where policy allows it, unreviewed output is not. The issue is verification, not the tool.
What to look for
- Checks what policy actually allows, then treats the unverified error as the real problem.
- Talks to the teammate directly before escalating.
- Proposes a check that would have caught it, not a ban.
Example answerRed flags
Example answer
I'd flag the error in review like any other error — that's what review is for. Then I'd talk to them privately: "the model wrote this and you didn't read it, right?" Not to shame them; I use AI drafts too. But whoever's name is on the work owns every claim in it, and this would've reached a client.
Red flags
- Where policy allows the tool, focuses on banning it rather than the missing verification step.
- Would fix the error silently and let the pattern continue.
You find an AI tool that would save you five hours a week, but using it means uploading client data, and company policy doesn't mention it. What do you do?
Good answers treat silence in policy as "ask," not "allowed." The strong candidate checks with whoever owns data or security before uploading anything, and looks for a way to get the gain without the exposure.
What to look for
- Defaults to asking, not to "nobody said I couldn't."
- Identifies the actual risk: client data leaving approved systems.
- Explores a middle path — anonymized data, an approved alternative.
Example answerRed flags
Example answer
Five hours a week is worth pursuing, so I wouldn't drop it — but I wouldn't upload client data on my own judgment either. I'd ask whoever owns security or the client contract, and meanwhile test the tool with dummy data to prove the savings. If the answer is no, the demo makes the case for an approved equivalent.
Red flags
- Would start using it and stop if someone objects.
- Doesn't register client data as the risky part of the scenario.
An AI assistant gives you an answer that contradicts what a senior colleague told you last week. The work is due today. How do you resolve it?
Neither source wins by authority. A strong answer goes to a primary source — documentation, the data itself, a quick test — and treats the deadline as a reason to check faster, not to skip checking.
What to look for
- Checks a primary source instead of picking whichever voice feels safer.
- Tells the colleague what the model said — as a question, not a gotcha.
- Ships with the discrepancy resolved or explicitly flagged, never buried.
Example answerRed flags
Example answer
Contradiction means at least one of them is wrong, and I can't know which by vibes. I'd spend thirty minutes on the primary source — the docs, the actual numbers. If that settles it, done; I'd mention it to the colleague either way. If I can't verify in time, I ship the safer interpretation with the open question flagged.
Red flags
- Automatically trusts the human, or automatically trusts the model.
- Ships whichever answer is more convenient and hopes.
A client asks whether AI was used to produce the deliverable you just sent. It was, partly. What do you say?
The only strong answer is the honest one, delivered without panic: what the tools did, what humans reviewed, and who is accountable. Hedging turns a workflow question into a trust problem.
What to look for
- Answers truthfully, without treating the question as an accusation.
- Can describe the human review step concretely.
- Owns accountability for the result regardless of what drafted it.
Example answerRed flags
Example answer
I'd tell them exactly what happened: "Yes — a model drafted sections two and three, and I verified every figure against your data myself." Then I'd ask what's behind the question. Sometimes it's a compliance requirement we should know about; sometimes they just want to hear that we stand behind the work.
Red flags
- Would deny it, or answer so vaguely the denial is implied.
- Sees no reason a client might legitimately need to know.
Questions 10–13
Situational questions about people and priorities
A teammate you depend on is visibly struggling and your shared deadline is at risk. They haven't asked for help. What would you do?
Strong answers open with the person, not the deadline — a private, specific offer of help — while quietly protecting the delivery. Going to the manager first, or ignoring it, both fail.
What to look for
- Approaches the teammate privately before doing anything else.
- Offers something specific — taking a task, pairing on a blocker — not "let me know if you need anything."
- Has a fallback for the deadline that doesn't humiliate the teammate.
Example answerRed flags
Example answer
I'd ask them for a walk, away from the task board: "You seem underwater — what can I take?" Specific offers get accepted; vague ones get declined. If the deadline still looks shaky after that, I'd suggest we flag it together, so it's our slip to manage and not a secret I'm keeping about them.
Red flags
- Reports the struggling teammate to the manager as the first move.
- Protects the deadline by silently redoing their work behind their back.
In a meeting, a colleague presents your idea as their own and it lands well. What do you do in the moment, and afterwards?
Good answers stay composed in the room — often adding a detail that signals authorship — and have the direct conversation afterwards. The goal is fixing the pattern, not winning the meeting.
What to look for
- In the moment: builds on the idea in a way that shows deep ownership, without a scene.
- Afterwards: talks to the colleague directly before involving anyone else.
- Distinguishes a one-off from a pattern and responds proportionally.
Example answerRed flags
Example answer
In the room I'd jump in and extend it — "right, and when I sketched this last week the tricky part was the rollout" — which recalibrates things without a confrontation. Afterwards, one direct question: "That was my proposal from Tuesday — what happened there?" Once is often carelessness. Twice is a pattern, and then it goes wider.
Red flags
- Would publicly correct the colleague mid-meeting to reclaim credit.
- Would say nothing, ever, and let resentment accumulate.
You inherit a task from someone who left the company. Their notes make no sense and it's due in three days. What's your approach?
Strong candidates timebox the archaeology — an hour on notes, files, and anyone who worked nearby — then rebuild from the task's purpose and renegotiate scope early if three days won't cover it.
What to look for
- Hunts for context beyond the notes: old threads, adjacent teammates, the requester.
- Works backwards from what the task is for, not from deciphering every line.
- Flags a scope or deadline problem on day one, not day three.
Example answerRed flags
Example answer
Hour one: skim the notes, search their old messages, ask whoever sat closest to the work. Then I'd go to whoever needs the output and ask what decision it feeds — that usually matters more than the notes did. If what they need doesn't fit three days, they hear that the same afternoon, with what I can deliver.
Red flags
- Spends two of the three days trying to decode the notes alone.
- Never contacts the person waiting on the output.
Your manager hands you a goal you privately believe is impossible in the time given. What do you do in the first day?
The strong move is surfacing the doubt immediately, with math: here's the gap, here are the options — more time, less scope, more hands. Accepting silently and failing later is the worst path.
What to look for
- Raises the concern on day one, framed as math rather than complaint.
- Comes with options attached: scope cuts, phasing, added resources.
- Stays open to being wrong — asks what the manager knows that they don't.
Example answerRed flags
Example answer
Day one I'd build the quickest honest estimate — even rough, it turns "this feels impossible" into "this is a six-week job in a three-week window." Then I'd bring it to my manager with choices: cut these two features, move the date, or add a person. Maybe she knows something that changes my math. If not, we caught it early.
Red flags
- Accepts the goal enthusiastically and plans to raise problems "if they come up."
- Declares it impossible without doing any estimation first.
Scoring rubric
| Score | Evidence anchor |
|---|---|
| 1 | Answers every scenario with a platitude — 'I'd communicate openly' — and no follow-up produces a first step, a sequence, or a named trade-off. |
| 2 | Gives textbook answers that would fit any company: plausible steps, no clarifying questions, no sense of what the plan costs or where it could go wrong. |
| 3 | Builds a reasonable plan with a clear first move and asks at least one clarifying question, but misses second-order effects — who else is affected, what the fix costs. |
| 4 | Handles hypotheticals like real work: asks what they'd need to know, sequences their actions, names the trade-off in each choice, and adapts when you add a complication. |
| 5 | Interrogates the scenario before answering, states assumptions out loud, and gives a decision with its trigger points — 'if X happens, I switch to Y.' The answers sound like recalled experience, not theory, and survive every complication you add. |
Frequently asked questions
What are situational interview questions?
Situational interview questions describe a hypothetical work scenario — "what would you do if…" — and ask the candidate to reason through a response. They test judgment and problem-solving on situations the candidate may never have faced.
Situational vs behavioral interview questions — which should you ask?
Behavioral questions look backward at what a candidate did; situational questions look forward at what they would do. In the behavioral vs situational choice, use situational where experience can't exist, and pair it with a real-example follow-up everywhere else.
Do these situational interview questions examples work across roles?
The 13 above are deliberately role-agnostic scenarios about deadlines, errors, clients, and AI — usable for most office roles as written. For a specific role, don't just rename the props: the role pages in this library use dilemmas, evidence, and scoring anchors built for that job's actual failure modes.
Are situational based interview questions the same as scenario-based questions?
Yes. Situational based interview questions, scenario questions, and hypothetical questions all name the same format: a described work situation plus "what would you do." STAR-style behavioral prompts are the different one — they require a real past event.
How should candidates answer situational interview questions?
Candidates are coached to give a structured plan: assess, act, communicate. Expect that polish and probe past it — add a complication mid-answer, then ask when they last faced something similar. Kira's first-round voice interviews push these follow-ups.