Design a structured interview loop for a specific role
advancedClaude OpusHR & PeopleRecruitingrecruitinginterview-designhiring-loopstructured-interviewtalent
Use case
Use this when opening a new role or rebuilding a loop that's leaking strong candidates or letting weak ones through. The most common failure mode is interviewers all measuring the same thing. This prompt forces you to map signals to stages and avoid that overlap.
The prompt
You are a Head of Talent who has designed hiring loops for Stripe, Airbnb, and a Series B startup that scaled from 30 to 300. Design a structured interview loop for the following role. Role context: - Title and level:{{title_level}}- Function and team:{{function_team}}- Reports to:{{reports_to}}- Top 3 outcomes the hire must drive in year one:{{key_outcomes}}- Team size and composition:{{team_context}}- Hiring volume (single hire vs. cohort):{{volume}}- Current candidate pipeline strength:{{pipeline}}(strong/medium/thin) Design a loop with: 1. **Signal map** — A table of the 5–7 signals you must assess (e.g., technical depth, judgment, ownership, communication, culture-add). For each signal, rate how easy it is to fake (1–5) and which stage best surfaces it. 2. **Stage-by-stage design (4–6 stages)** — For each stage: - Stage name and length - Interviewer role (be specific — "EM, not the hiring manager" matters) - Primary signals assessed (from the signal map) - Format (behavioral, work sample, case, panel, etc.) - 2–3 example questions or prompts - Pass/fail bar in plain language - What disqualifies — make this concrete 3. **Calibration plan** — How interviewers align before the loop starts. Include a 30-min calibration meeting agenda. 4. **Debrief format** — How the hire/no-hire decision actually gets made. Include who has veto power and how to break ties. 5. **Time-to-decision target** — From first conversation to offer, in business days. Most loops are too slow; say what speed is realistic for this role. 6. **Likely failure modes** — Where this loop will leak good candidates or let bad ones through, and how to mitigate. Avoid: trivia questions, brain teasers, "tell me about yourself" as an opener for late-stage interviews, redundant signals across stages, panels that don't have a clear scribe.
Variables
{{{{title_level}}}}Replace with your {{title level}}{{{{function_team}}}}Replace with your {{function team}}{{{{reports_to}}}}Replace with your {{reports to}}{{{{key_outcomes}}}}Replace with your {{key outcomes}}{{{{team_context}}}}Replace with your {{team context}}{{{{volume}}}}Replace with your {{volume}}{{{{pipeline}}}}Replace with your {{pipeline}}
Example
Input
title_level: Senior Product Manager (IC-5) function_team: Growth Product, owns activation and onboarding for self-serve SMB users reports_to: Director of Growth Product key_outcomes: (1) Lift activation rate from 28% to 35% within 9 months. (2) Ship and validate 3 onboarding experiments per quarter. (3) Reduce time-to-first-value from 14 days to 7. team_context: Team of 1 PM (vacant — this hire), 4 engineers, 1 designer, 0.5 data scientist. Director hands-on but stretched across 3 PMs. volume: Single hire, urgent — slot has been open 11 weeks pipeline: Medium. Strong inbound from past Stripe/Square candidates, weaker on healthcare/regulated SaaS.
Output
# Interview Loop — Senior PM, Growth (Activation)
## 1. Signal map
| Signal | Why it matters here | Fake-ability (1–5) | Best stage to assess |
|---|---|---|---|
| Growth product judgment | Has to pick the right experiments from a long list | 4 | Case study (Stage 3) |
| Quantitative rigor | Activation work is metric-driven; can't survive without funnel literacy | 2 | Case study + EM technical |
| Ownership / drive | Stretched team needs a PM who runs, not asks | 5 | Behavioral with hiring manager + reference |
| Cross-functional leadership | Will lead 4 eng + designer with no formal authority | 4 | Panel with eng + design |
| Communication (written) | Growth experiments are won and lost in the doc | 3 | Take-home or pre-read |
| Self-serve / PLG instinct | Domain-specific; product-led intuition is hard to teach | 3 | Case study + portfolio walkthrough |
| Culture-add (humility, candor) | Director is stretched; this PM has to push back well | 5 | Skip-level + values interview |
Two key calls: ownership and culture-add are the hardest to fake-detect, so they get the most senior interviewers and are checked twice.
## 2. Stages
### Stage 1 — Recruiter screen (30 min)
- **Interviewer:** Recruiter
- **Signals:** Basic fit, comp alignment, motivation
- **Format:** Phone, conversational
- **Questions:** "Walk me through what's drawing you to growth roles right now." "What's your comp expectation and what's behind that number?" "What's your timeline?"
- **Pass bar:** Coherent narrative, comp within range or negotiable, real interest in self-serve growth
- **Disqualifies:** No PLG/self-serve experience whatsoever, comp 25%+ above band with no flex, looking primarily for IC-6 title
### Stage 2 — Hiring manager (45 min)
- **Interviewer:** Director of Growth Product
- **Signals:** Ownership, growth product judgment, communication
- **Format:** Behavioral with portfolio walkthrough
- **Questions:** "Tell me about an activation or onboarding experiment you ran end-to-end. Walk me through the funnel before, the hypothesis, what shipped, what you learned." "Describe a time you killed a project that was going to ship. Why and how?" "What's a metric you've moved that you're proud of, and one you tried to move and couldn't?"
- **Pass bar:** Specific numbers in answers, owns failures without deflection, can articulate why a specific experiment was the right next bet
- **Disqualifies:** Vague metrics ("we improved engagement"), blames team for failures, can't explain prioritization logic
### Stage 3 — Case study (60 min, sent 48h ahead)
- **Interviewer:** Two PMs from adjacent teams (not Growth)
- **Signals:** Growth judgment, quantitative rigor, PLG instinct
- **Format:** Pre-read: anonymized funnel data + 3 candidate experiment ideas. Live: candidate presents which one they'd pick and why, then handles pushback.
- **Sample prompt:** "Here's our activation funnel. Drop-off is 22% at step 3. We have bandwidth to ship one of three experiments next quarter — A/B/C. Pick one, defend it. Then tell us what you'd kill if forced to deprioritize one current initiative."
- **Pass bar:** Anchors recommendation in funnel math, articulates a falsifiable hypothesis, handles pushback by updating not defending
- **Disqualifies:** Recommends all three, can't read the funnel, defends the wrong answer when challenged
### Stage 4 — Cross-functional panel (60 min, two interviewers)
- **Interviewers:** Tech lead from Growth eng + Senior designer from Growth
- **Signals:** Cross-functional leadership, communication, culture-add
- **Format:** Two scenarios + one collaborative whiteboarding exercise
- **Sample prompt:** "An engineer pushes back on your top experiment because it requires touching auth code. They estimate 4 weeks; you have 2. Walk us through how you handle the next 24 hours." Then collaborative: "Sketch the new onboarding flow with us."
- **Pass bar:** Listens, integrates input, doesn't fold under disagreement, treats eng/design as partners not stakeholders
- **Disqualifies:** Talks over interviewers, frames eng as a constraint to manage, can't sketch a flow
### Stage 5 — Skip-level + values (45 min)
- **Interviewer:** VP of Product
- **Signals:** Culture-add, ownership, judgment under ambiguity
- **Format:** Behavioral, deep
- **Questions:** "Tell me about a time you disagreed with your manager and didn't back down. What happened?" "What's something you believe about product that most PMs you know don't?" "When have you been wrong about a strategic call, and how did you find out?"
- **Pass bar:** Genuine candor, specific examples, owns being wrong with detail
- **Disqualifies:** Polished non-answers, can't name a real disagreement, no self-criticism
## 3. Calibration plan
Hold a 30-min calibration meeting before the first onsite, attended by all 6 interviewers. Agenda:
- 5 min: Director walks through the role scorecard and the year-one outcomes
- 10 min: Walk the signal map; each interviewer claims their assigned signals out loud
- 10 min: Calibration on a sample candidate (use a recent strong/weak case from another loop)
- 5 min: Confirm pass bars and disqualifiers; align on scoring scale (1–5, 3 is hire bar)
## 4. Debrief format
- Each interviewer submits written feedback in scorecard within 24h, before discussing with anyone else (avoids anchoring)
- 45-min synchronous debrief, hiring manager facilitates but speaks last
- Vote: hire / no-hire / strong-no-hire on a written scorecard before discussion
- **Veto:** Hiring manager can veto a hire-vote majority. No interviewer can veto a no-hire — but the hiring manager can override one strong-no if they have a specific theory why
- Tiebreaker: hiring manager's call, with rationale documented
## 5. Time-to-decision target
10–12 business days from first recruiter screen to offer, given the role has been open 11 weeks. Compress Stage 3 turnaround (48h pre-read → next-day live) and run Stages 4 and 5 on the same day.
## 6. Likely failure modes
- **Risk:** Stages 2 and 5 both probe ownership behaviorally — they may double-count the same signal. *Mitigation:* Stage 5 should focus on disagreement and being-wrong specifically, not general ownership stories.
- **Risk:** Case study favors candidates who've seen this exact funnel pattern before, biasing toward Stripe/Square alums. *Mitigation:* Score for reasoning quality not answer correctness; have one of the two case interviewers be from a different industry.
- **Risk:** With slot open 11 weeks, hiring manager may lower the bar. *Mitigation:* VP signs off on every offer; debrief explicitly asks "would we hire this person if the role had been open 2 weeks?"
- **Risk:** No data scientist in the loop, but the role is metric-heavy. *Mitigation:* Add the 0.5 DS as a 30-min informal chat in Stage 4 — not scored, but vetoes obvious quant gaps.
Tips for best results
- 1Map every signal to exactly one or two stages. If three interviewers are all measuring 'communication,' you're wasting two of them.
- 2Write disqualifiers in plain language before the loop starts. 'Vague metrics in answers' is a real disqualifier; 'culture fit' is not.
- 3The case study should reflect actual work the person will do in week one. If your case is generic 'design a feature for Spotify,' you're not learning what you need to learn.
- 4Always have feedback submitted in writing before the debrief discussion. Anchoring is the #1 reason hiring loops produce inconsistent decisions.
- 5Time-to-decision matters more than people admit. The best candidates have multiple loops running. A 4-week loop loses to a 10-day loop nine times out of ten.
Related prompts
Generate structured interview questions for a role
intermediateGenerate a bank of structured behavioral and situational interview questions for a specific role, organized by competency and designed to produce evidence-based hiring decisions.
HR & Peoplerecruitinginterviewstructured-hiring
Create an interview scorecard for a role
advancedGenerate a structured interview scorecard with evaluation dimensions, rating criteria, and scoring guidance to make hiring decisions more consistent and defensible.
HR & Peoplerecruitinginterviewscorecard
Generate reference check questions for a specific role
intermediateGenerate a structured set of reference check questions tailored to a specific role that go beyond confirming employment dates and surface real information about how a candidate performs.
HR & Peoplerecruitingreference-checkhiring
Need help implementing this prompt in your workflow?
Book a call