Mock Interview
When To Use
You have prepared on paper and never said any of it out loud. The stories are written, the case steps are known, and you have not yet heard how an answer sounds at minute three, with someone asking "why that option?". It answers: what happens to my preparation when a person interrupts it? Practice only, before the interview.
When Not To Use
Never during a real interview: no listening in, no live transcript, no answers while you are in the room or on the call. To learn the method of one round first, use Product Sense Case, Metrics Case, Prioritisation and Execution Case or App Critique. With no stories chosen for a behavioural round, run Behavioural Interview Answers first.
Inputs
- The round you want to rehearse, as it stands on your Interview Loop Map: format, length, the interviewer's role
- The prep for that round: your Behavioural Round Prep Sheet, story bank, case notes or walkthrough outline
- Your Career Evidence Inventory, so a sharper version can draw only on your real work
If you have none of this, I start from the role title and the round format you name and mark the output as a first draft.
Approach
Rehearsal role-play, the mock interview any careers service runs, with an answer rubric you can read before the first question. Question types follow the US Office of Personnel Management's description of structured interviews: predetermined questions, in a fixed order, about past behaviour or a hypothetical situation. The judgment: notes are about the answer, quoted in your own words, and the rubric is a stand-in built from public guides, not any employer's scoring sheet. The failure it prevents is the answer that reads well on the page and runs four minutes before reaching what you did.
Workflow
- Ask at most three questions: which round (recruiter screen, behavioural or values, product sense, metrics, execution, app critique, walkthrough questions, stakeholder or triad round), how long, and notes after each answer or at the end.
- Set the stand-in: a generic interviewer for that round and role, built only from your loop map and sourced notes. I say plainly that it is a stand-in, never the real interviewer, and I attribute no opinion or quote to a real person.
- Show the rubric before the round starts. Behavioural rows: situation brief, your task, your actions as "I", result with what you learned, answered the question asked, length. A case round uses the rubric rows of the matching case skill in this pack. Recruiter screen rows: why you are looking in two true sentences, timing, pay handled as you planned, your questions asked.
- Run the round in its real order: a short opening and introduction; the main questions one at a time; two or three follow-ups per answer (what was your part, why that option, what if the constraint changes); your own questions; a close on time. I do not help mid-answer. A stall is part of the practice. Voice mode lets you say it aloud and hear the follow-ups; text works too.
- Write the notes: one table per answer, each rubric row marked present, thin or missing, with your own words quoted as the evidence. No number, no total, no verdict on you.
- Rebuild the weakest answer once, using only facts you gave in the round or rows in your inventory, with any gap left as
[placeholder]. Then list what to practise again.
- Close every run with the same line: practice only; hold the real interview yourself, without Claude.
Output Format
# Mock Interview Notes
**Round:** [format] · **Stand-in for:** a [role] interviewer (a stand-in, not the real person) · **Length:** [minutes] · **Mode:** [voice / text] · **Date:** [date]
## Rubric shown before the round
| Row | What is listened for |
|---|---|
| [row] | [description] |
## Answer [n]: [question as asked]
| Rubric row | Status (present / thin / missing) | Your words |
|---|---|---|
| [row] | [status] | "[quoted from your answer]" |
**Follow-ups asked:** [follow-up] · [follow-up]
## Sharper version of the weakest answer
**Question:** [question] · **What you said:** "[short quote]"
**Sharper, from your facts only:** [rebuilt answer, with [placeholder] where you gave no fact]
## What to practise again
- [One part of one answer, and how to practise it]
**Reminder:** practice only. You hold the real interview yourself, without Claude.
## Decision
You decide what to change in your stories or notes and whether to run the round again, by [date, before the real round].
Done When
- The rubric was shown before the first question, and the round ran in its real order and closed on time, including your own questions.
- Every answer has a table with each row marked present, thin or missing and your words quoted.
- The sharper version contains no fact you did not give, and the notes end with the reminder.
Quality Bar
- Notes describe the answer: no number, grade, rank, hire or no-hire call, and no comparison with other candidates.
- The stand-in never gets views or quotes from a real interviewer, and it does not coach while you are answering.
- No model answer to memorise: the sharper version shows shape, in your facts, for you to say your own way.
- A recording of your own practice is yours to keep. Recording a real interview raises consent and legal questions: check with a qualified adviser.
- Practice only: Claude is never used during a real interview, and the sharper version uses only facts you gave.
Next
Run pmjob-ai-use-plan (Candidate AI Use Plan) to settle what AI use each employer allows.
About the makers
This pack is made by 90n (90n.ai): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).
1---2name: pmjob-mock-interview3description: Runs a practice interview round where Claude plays a stand-in interviewer for the format you choose and asks follow-ups, then writes Mock Interview Notes on each answer against a rubric you see first, as present, thin or missing, with one sharper version of your weakest answer in your own facts. Use for "run pmjob-mock-interview", "mock product interview", "practise my interview out loud", "play the interviewer", "rehearse the behavioural round", "practice interview with follow-up questions", "notes on my interview answers", "rehearse before the final round", part of the Claude Guide for Product Interviews Pack by 90n.4metadata:5 source: https://skillmd.com/skills/polar-bear-org/pmjob-mock-interview6---78# Mock Interview910## When To Use11You have prepared on paper and never said any of it out loud. The stories are written, the case steps are known, and you have not yet heard how an answer sounds at minute three, with someone asking "why that option?". It answers: what happens to my preparation when a person interrupts it? Practice only, before the interview.1213## When Not To Use14Never during a real interview: no listening in, no live transcript, no answers while you are in the room or on the call. To learn the method of one round first, use Product Sense Case, Metrics Case, Prioritisation and Execution Case or App Critique. With no stories chosen for a behavioural round, run Behavioural Interview Answers first.1516## Inputs17- The round you want to rehearse, as it stands on your Interview Loop Map: format, length, the interviewer's role18- The prep for that round: your Behavioural Round Prep Sheet, story bank, case notes or walkthrough outline19- Your Career Evidence Inventory, so a sharper version can draw only on your real work20If you have none of this, I start from the role title and the round format you name and mark the output as a first draft.2122## Approach23Rehearsal role-play, the mock interview any careers service runs, with an answer rubric you can read before the first question. Question types follow the US Office of Personnel Management's description of structured interviews: predetermined questions, in a fixed order, about past behaviour or a hypothetical situation. The judgment: notes are about the answer, quoted in your own words, and the rubric is a stand-in built from public guides, not any employer's scoring sheet. The failure it prevents is the answer that reads well on the page and runs four minutes before reaching what you did.2425## Workflow261. Ask at most three questions: which round (recruiter screen, behavioural or values, product sense, metrics, execution, app critique, walkthrough questions, stakeholder or triad round), how long, and notes after each answer or at the end.272. Set the stand-in: a generic interviewer for that round and role, built only from your loop map and sourced notes. I say plainly that it is a stand-in, never the real interviewer, and I attribute no opinion or quote to a real person.283. Show the rubric before the round starts. Behavioural rows: situation brief, your task, your actions as "I", result with what you learned, answered the question asked, length. A case round uses the rubric rows of the matching case skill in this pack. Recruiter screen rows: why you are looking in two true sentences, timing, pay handled as you planned, your questions asked.294. Run the round in its real order: a short opening and introduction; the main questions one at a time; two or three follow-ups per answer (what was your part, why that option, what if the constraint changes); your own questions; a close on time. I do not help mid-answer. A stall is part of the practice. Voice mode lets you say it aloud and hear the follow-ups; text works too.305. Write the notes: one table per answer, each rubric row marked present, thin or missing, with your own words quoted as the evidence. No number, no total, no verdict on you.316. Rebuild the weakest answer once, using only facts you gave in the round or rows in your inventory, with any gap left as `[placeholder]`. Then list what to practise again.327. Close every run with the same line: practice only; hold the real interview yourself, without Claude.3334## Output Format35```markdown36# Mock Interview Notes37**Round:** [format] · **Stand-in for:** a [role] interviewer (a stand-in, not the real person) · **Length:** [minutes] · **Mode:** [voice / text] · **Date:** [date]38## Rubric shown before the round39| Row | What is listened for |40|---|---|41| [row] | [description] |42## Answer [n]: [question as asked]43| Rubric row | Status (present / thin / missing) | Your words |44|---|---|---|45| [row] | [status] | "[quoted from your answer]" |46**Follow-ups asked:** [follow-up] · [follow-up]47## Sharper version of the weakest answer48**Question:** [question] · **What you said:** "[short quote]"49**Sharper, from your facts only:** [rebuilt answer, with [placeholder] where you gave no fact]50## What to practise again51- [One part of one answer, and how to practise it]52**Reminder:** practice only. You hold the real interview yourself, without Claude.53## Decision54You decide what to change in your stories or notes and whether to run the round again, by [date, before the real round].55```5657## Done When58- The rubric was shown before the first question, and the round ran in its real order and closed on time, including your own questions.59- Every answer has a table with each row marked present, thin or missing and your words quoted.60- The sharper version contains no fact you did not give, and the notes end with the reminder.6162## Quality Bar63- Notes describe the answer: no number, grade, rank, hire or no-hire call, and no comparison with other candidates.64- The stand-in never gets views or quotes from a real interviewer, and it does not coach while you are answering.65- No model answer to memorise: the sharper version shows shape, in your facts, for you to say your own way.66- A recording of your own practice is yours to keep. Recording a real interview raises consent and legal questions: check with a qualified adviser.67- Practice only: Claude is never used during a real interview, and the sharper version uses only facts you gave.6869## Next70Run pmjob-ai-use-plan (Candidate AI Use Plan) to settle what AI use each employer allows.7172## About the makers7374This pack is made by 90n ([90n.ai](https://90n.ai)): AI-native ways of working and tools for Product, Marketing and HR teams, built by ex-McKinsey founders. We map where time and money leak, redesign the work around AI and stay until your people run it themselves, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).