Behavioural interview prep

The STAR framework

Every behavioural answer should have the same four beats, in order. Each beat has a specific length and a specific job.

1. 1

   Situation

   One-sentence setup. Where, when, what the team was doing. Keep it short — the interviewer wants the Task fast.

2. 2

   Task

   What was yours to own. The specific problem assigned to you, not the team's broad goal. Framed in first person: 'I was the lead on …'.

3. 3

   Action

   What you did — step by step, decisions made, trade-offs considered, teammates coordinated with. This is the longest part and the one that reveals judgement.

4. 4

   Result

   The outcome with a number if there is one — latency saved, bug count dropped, people onboarded. If there isn't a number, be honest about why: 'We shipped it, retention didn't move; here's what I think that meant.'

15 questions you'll be asked

Pre-write one story per question. Rehearse each to under two minutes. A pre-rehearsed story that is still your real story reads as preparation, not performance — the opposite is what fails.

1. 1

   Tell me about a time you disagreed with a teammate.

   How to answer

   Describe the specific disagreement — a technical choice is better than an interpersonal one. Walk through how you presented your view, how you listened to theirs, what the resolution process looked like (not a vote, not a title fight — a conversation grounded in evidence), and what you learned about the other person's reasoning.

   Common mistakes

   • Framing yourself as right and the other person as wrong.
   • A story where the resolution was 'they came around to my view' with no acknowledgement of their valid points.
   • Zero concrete detail — 'we talked it out' is not an answer.

   Signal

   The signal is intellectual humility paired with the willingness to make a case. Both halves matter.

2. 2

   Tell me about a time you made a mistake.

   How to answer

   A real mistake with a real consequence — not 'I worked too hard'. Describe what you missed, how it was caught, what the direct cost was, and what you specifically changed in your own process afterwards. The last part is the actual signal.

   Common mistakes

   • A mistake that is secretly a strength ('I was too thorough').
   • No real consequence — the story feels invented.
   • No lasting change in your process — 'I'll be more careful next time' is not an answer.

   Signal

   The absence of a credible failure story reads as insufficient scar tissue, especially at Staff+.

3. 3

   Tell me about a time you had to make a hard call under ambiguity.

   How to answer

   The specific call, the information you had and the information you didn't, how you decided between the real options, what you did to make it recoverable in case you were wrong (reversibility is a strong frame), what the outcome was.

   Common mistakes

   • A story where the right answer was obvious in hindsight.
   • No framing of reversibility or risk-mitigation.
   • Blame framed outward — 'we had to decide because PM was indecisive'.

   Signal

   Senior+ loops reward 'I chose, I made it reversible, I was ready to be wrong' far more than 'I chose and I was right'.

4. 4

   Tell me about a time you influenced someone without authority.

   How to answer

   The person or team, what you were trying to change, how you built the case (writing an RFC, sharing data, prototyping), how you handled pushback, what the outcome was. Emphasise listening and iterating — not persuasion as a one-way broadcast.

   Common mistakes

   • A story that is really 'I complained until they agreed'.
   • No concrete artefact (doc, prototype, data) you produced.
   • The influence came from title or authority after all.

   Signal

   At Senior+ the influence-without-authority story is one of the three or four interviewers probe most.

5. 5

   Tell me about a project you're most proud of.

   How to answer

   Pick a project where your specific contribution was material and the outcome is measurable. Describe the problem, the architecture or decision set, your role specifically, what was hard, the measurable outcome. End with what you'd do differently with hindsight.

   Common mistakes

   • Pride tied to team outcomes without clarity on your part.
   • No measurable outcome — 'users seemed to love it'.
   • No 'what I'd do differently' — Senior+ reads this as insufficient reflection.

   Signal

   The 'differently with hindsight' addition is a signal multiplier at every level.

6. 6

   Tell me about a time you had to deliver bad news.

   How to answer

   The context, who you had to tell, what you told them, how you handled questions (ideally including the ones you couldn't fully answer), and what happened next.

   Common mistakes

   • The bad news is really someone else's fault.
   • You delegated the delivery.
   • No follow-up — what you did after the conversation.

   Signal

   The signal is composure and ownership. Both.

7. 7

   Tell me about a time you had to work with someone difficult.

   How to answer

   Describe the specific behaviour (not the person) that was difficult, how you adapted your approach, what you learned about yourself from the interaction, and how the working relationship ended up.

   Common mistakes

   • Characterising the other person as 'toxic' or 'incompetent'.
   • No self-awareness about your contribution to the dynamic.
   • A story where you avoided the person rather than engaging.

   Signal

   Interviewers are listening for whether you'd be the difficult person for a future teammate's story.

8. 8

   Tell me about a time you took initiative on something nobody asked for.

   How to answer

   The problem you noticed, why you were the right person to notice it, how you validated it was worth doing (not jumping to solutions), what you built or proposed, how you got buy-in, the outcome.

   Common mistakes

   • The initiative was really assigned work with light framing.
   • No validation step — you built first, convinced later.
   • The outcome is 'and now the team uses it' with no number.

   Signal

   Ownership stories are strongest when they include a validation step before action.

9. 9

   Tell me about a time you mentored someone.

   How to answer

   Who, what they were working on, how you adapted your approach to their level and learning style, how you measured progress (not 'they seemed happy'), what they can do now that they couldn't before.

   Common mistakes

   • Mentoring described as 'answered their questions'.
   • No adaptation to their style — your mentoring is one-size-fits-all.
   • No measurable change in their capability.

   Signal

   Senior+ loops explicitly probe for repeatable mentorship, not isolated help.

10. 10

    Tell me about a time you missed a deadline.

    How to answer

    The deadline, the slippage you noticed early (because you noticed, right?), how you communicated it, what you changed in-flight to recover scope, what you did afterwards to prevent the same cause next time.

    Common mistakes

    • Blame on dependencies with no self-ownership.
    • No early communication — you noticed the slip but didn't flag it until it was late.
    • No process change afterwards.

    Signal

    The 'how I surfaced it early' part is the biggest signal — the actual slip less so.

11. 11

    Tell me about a time you changed your mind on a technical opinion.

    How to answer

    The opinion you held, what caused you to reconsider (new data, a mentor's pushback, a bug report), the specific thing you updated, and what you now hold more loosely because of it.

    Common mistakes

    • A trivial opinion change — 'I used to prefer tabs, now spaces'.
    • A change forced by authority, not by evidence.
    • No reflection on what else you now hold less firmly.

    Signal

    This question is the shortcut to assessing intellectual honesty. Prepare one real example.

12. 12

    Tell me about a time you had to prioritise.

    How to answer

    The competing demands, the explicit criteria you used (impact, reversibility, cost), the choice you made, how you communicated what was being deprioritised, the outcome.

    Common mistakes

    • No explicit criteria — the choice reads as taste.
    • No communication about what was dropped — it just quietly disappeared.
    • Everything was equally important; you worked harder to do it all.

    Signal

    Explicit criteria and visible communication are both the signal here.

13. 13

    Tell me about a time you had to deal with ambiguity in requirements.

    How to answer

    The ambiguity, how you spotted it (because you were asking questions), who you went to for clarification, how you proceeded when you still didn't have a crisp answer, the outcome.

    Common mistakes

    • You just built what you thought was intended and shipped.
    • You waited for a crisp spec and did nothing in the meantime.
    • No lesson about what you'd ask earlier next time.

    Signal

    Senior+ loops specifically probe for this because the day-job is ambiguity. Prepare one real example.

14. 14

    Tell me about a time you had to say no.

    How to answer

    The request, why saying yes would have been wrong (not just hard — wrong), how you framed the no, what alternative you offered or helped find, how the relationship held up.

    Common mistakes

    • The 'no' was really a 'yes, but later'.
    • The 'no' broke the relationship.
    • No alternative offered — just a refusal.

    Signal

    The alternative-offered part is what separates 'can say no' from 'is an obstacle'.

15. 15

    What's a strongly-held belief about engineering you've genuinely tested?

    How to answer

    A real opinion (microservices-vs-monolith, test coverage, TypeScript migrations, you pick), the evidence you've gathered for it, the counter-evidence you take seriously, and the specific conditions under which you'd update it.

    Common mistakes

    • A pop-tech-twitter opinion restated as yours.
    • No counter-evidence named — the belief is held unexamined.
    • No update conditions — the belief is non-falsifiable.

    Signal

    At Staff+ this question is often the deepest filter. Prepare one real belief.

Map your stories to company value frameworks

Most companies have an explicit values-shaped behavioural rubric. Pre-mapping each of your stories to a specific value or two from the target company's framework is the single highest-leverage prep.

• Amazon

  16 Leadership Principles — Customer Obsession, Ownership, Invent and Simplify, Are Right A Lot, Learn and Be Curious, Hire and Develop the Best, Insist on the Highest Standards, Think Big, Bias for Action, Frugality, Earn Trust, Dive Deep, Have Backbone / Disagree and Commit, Deliver Results, Strive to be Earth's Best Employer, Success and Scale Bring Broad Responsibility.

  Full guide →

• Google

  Googleyness and Leadership — comfort with ambiguity, bias to action, intellectual humility, collaboration, and 'making the world's information useful' as the mission backdrop.

  Full guide →

• Microsoft

  Growth Mindset — learning from feedback, pivoting when wrong, demonstrating curiosity, admitting what you don't know. The Satya-era culture is explicit about these as scored signals.

  Full guide →

• Meta

  'Impact', 'Conflict', 'Hardest Project', 'Mistake' — the Jedi behavioural round is explicit. Measurable outcomes in each story.

  Full guide →

• Atlassian

  Five values — Open company, no bullshit; Build with heart and balance; Don't #@!% the customer; Play as a team; Be the change you seek.

  Full guide →

• GitLab

  CREDIT values — Collaboration, Results, Efficiency, Diversity-Inclusion-Belonging, Iteration, Transparency. Plus explicit remote-collaboration signals.

  Full guide →

• Netflix

  Culture Deck specifics — Freedom and Responsibility, Informed Captain, Keeper Test, Context Not Control, High Trust / Low Rules.

  Full guide →

• Stripe

  Written-sample follow-up; 'Sell Stripe to Stripe' as the product-reasoning round. Correctness, craft, writing clarity.

  Full guide →

• Airbnb

  Six core values — Be a host; Be a cereal entrepreneur; Champion the mission; Embrace the adventure; Simplify; Have the humility to serve. Product-first framing throughout.

  Full guide →

• Apple

  Discretion, craftsmanship, team-fit. Less rubric-explicit than most; prepare stories that credit specific design or UX decisions.

  Full guide →

• Anthropic

  Mission alignment, intellectual honesty, comfort with uncertainty. The behavioural round rewards nuance over confident one-liners.

  Full guide →

• Nvidia

  Technical depth, performance obsession, written clarity. Behavioural stories grounded in specific technical decisions carry more weight than generic ones.

  Full guide →

On the day

• Say the value or principle out loud. 'This is a Customer Obsession story. The situation was…' — explicit mapping lands cleaner than hoping the interviewer notices.
• Keep stories to two minutes. If you need to go longer, name why: 'The context matters here, so I'll take an extra minute.'
• Numbers wherever numbers exist. 'Dropped TTI by 400ms' beats 'noticeably faster'. If there's no number, say why.
• Lead with 'I', not 'we'. The behavioural round is scoring you, not your team. 'We shipped' is not an answer.
• End every story with what you learned or would do differently. The 'differently with hindsight' close is a signal multiplier at every level.
• Have a 'tell me about yourself' answer ready that is one minute, specific, and ends with a forward-looking statement about why this role.
• Prepare one real question per interviewer about their team's current hardest problem. Generic questions read as uninvested.