Skip to content

borghei/Claude-Skills / project-management/career/pm-interview-prep

pm-interview-prep

Verified

Plans PM interview prep across product sense, execution, strategy, behavioral and technical rounds, calibrated to level. Use when a PM loop is scheduled, you are running mock interviews, or you are moving up a level.

references/red-flags.md

View on GitHub
# Red Flags: PM Interview Prep

> Common ways this skill's output goes wrong — concrete examples, why they're bad, and how to fix them. Pair with the SKILL.md and Troubleshooting table.

## How to use this document

When you have just drafted answer outlines, completed a mock interview, or finished prepping, scan the red flags below before your real loop. Each red flag shows a *bad* and *good* version of an answer or behavior.

---

## Red Flag 1: Rote Framework Recitation

**Symptom.** Candidate opens the product-sense answer with "I'll use CIRCLES: Comprehend, Identify, Report, Cut, List, Evaluate, Summarize." Then mechanically walks through each letter.

**Why it's bad.** Frameworks are scaffolding, not the answer. Interviewers want to see *thinking*; reciting CIRCLES letter by letter signals you have memorized a template but cannot adapt it. Strong candidates use frameworks invisibly — the interviewer feels the structure without hearing the acronym.

**Bad example:**
> "For this question I'll use the CIRCLES Method. First, C — comprehend. What's the situation? It's a ride-sharing app..."

**Good example:**
> "Let me clarify a few things first — when you say 'redesign', is the goal to improve a specific metric, or open-ended? And who are the priority users — drivers or riders? Once we know that, I'll narrow to a user segment and define the job to be done."

**How to catch it.** Record yourself doing a mock. If you say the framework name out loud, you are leaning on the scaffolding too visibly.

---

## Red Flag 2: Ignoring Interviewer Signals (Talking Past Hints)

**Symptom.** Interviewer says "let's focus on monetization" and you continue brainstorming features. Or interviewer says "you've covered users well, let's move to metrics" and you keep listing user segments.

**Why it's bad.** Interviewers actively steer the conversation toward the dimensions they want to assess. Ignoring the steer signals poor listening and poor stakeholder skills — both red flags for a PM hire.

**Bad example:**
> Interviewer: "Let's pick one feature and go deep on metrics." Candidate: "Sure — and also let me list two more features I considered..."

**Good example:**
> Interviewer: "Let's pick one feature and go deep on metrics." Candidate: "Got it. Picking the in-ride safety feature. For metrics I'd start with leading — incidents reported per ride, response time. Counter-metrics — false-positive reports, driver satisfaction."

**How to catch it.** After any interviewer interjection, the next 30 seconds of your answer should address what they said. If you continue your previous thread, you missed the steer.

---

## Red Flag 3: Behavioral Answers Without Concrete Impact (Missing the "R" in STAR)

**Symptom.** Behavioral answer covers Situation, Task, and Action in detail. The Result is one vague sentence: "It went well and we learned a lot."

**Why it's bad.** Without a measurable result, the story is a process report, not an impact story. Interviewers cannot calibrate the candidate's level without seeing what changed. The "Result" is where seniority signal lives.

**Bad example:**
> "...we shipped the redesign in 6 weeks. It was well received. We learned the importance of user research."

**Good example:**
> "...we shipped the redesign in 6 weeks. Activation rose from 38% to 51% (measured 30-day cohort). NPS for new users moved from 18 to 34. The retro identified that we should have done one more usability test — I now build that into every launch."

**How to catch it.** For every behavioral answer, can you state the Result in numbers + a learning? If not, the story is incomplete.

---

## Red Flag 4: Stating Opinions as Facts in Product Sense

**Symptom.** Candidate asserts user behaviors ("users always want X") or market truths ("the average user spends 20 minutes per day") without sourcing or hedging.

**Why it's bad.** PMs who confuse hypothesis with fact alarm interviewers. The PM job is to *generate testable beliefs*, not to assert truth. Confident assertions without sourcing signal poor product judgment.

**Bad example:**
> "Riders want faster pickups more than anything. That's the #1 driver of NPS."

**Good example:**
> "My hypothesis is that pickup time is the strongest NPS driver for riders. To validate, I'd look at correlation in NPS data segmented by pickup time, and test with a survey. If wrong, I'd next check pricing and driver behavior."

**How to catch it.** If you have not said "my hypothesis", "I'd validate by", or "evidence would be" in the answer, you are likely asserting.

---

## Red Flag 5: Estimation Answers Without Showing the Math

**Symptom.** Candidate jumps to a final number ("there are about 50 million users") without showing the breakdown.

**Why it's bad.** Estimation interviews assess *structured thinking*, not the number itself. An answer without visible breakdown gives the interviewer nothing to evaluate. Even a wrong number with good math beats a right number from thin air.

**Bad example:**
> "I'd estimate there are 5 million weekly active users for this."

**Good example:**
> "Population US 330M, age 18-65 ~60% = 200M. Smartphone penetration ~80% = 160M. Category penetration ~30% = 48M. Of those, weekly-active ~20% = 9.6M. I'd sanity-check against published data on category — does that order of magnitude feel right?"

**How to catch it.** Re-read your answer. Are the multipliers visible? If not, you skipped the math.

---

## Red Flag 6: Behavioral Stories That Make You the Sole Hero

**Symptom.** Every behavioral story is "I did X, I drove Y, I decided Z" — no acknowledgment of the team, the cross-functional partners, or the dependencies.

**Why it's bad.** PMs work through influence. A story where the PM did everything alone reads as either (a) inflated, or (b) genuinely worked alone — which signals weak collaboration. Both interpretations damage the case.

**Bad example:**
> "I single-handedly turned the project around. I designed the solution, I aligned stakeholders, I drove engineering."

**Good example:**
> "The project was off-track. I partnered with the EM to scope down — she identified the biggest engineering risks; I owned the cust-facing trade-offs. Together we re-baselined with stakeholders. My specific contribution was the stakeholder framing memo and the weekly tracking ritual."

**How to catch it.** Count "I" vs "we" in your answer. If "I" outnumbers "we" 5-to-1, you sound like a lone wolf.

---

## Red Flag 7: Strategy Answer with No Trade-Off

**Symptom.** Candidate proposes a strategy and lists 5 reasons it is great. No discussion of what is being given up, what alternatives were considered, or what could go wrong.

**Why it's bad.** Strategy *is* trade-offs. A strategy answer with no trade-off is a wish list. Senior interviewers reject this because real strategic work requires saying no.

**Bad example:**
> "We should go upmarket. It's higher LTV, better margins, less churn, more defensible, and TAM is large."

**Good example:**
> "We should go upmarket. Trade-offs: longer sales cycles (12-18 months vs 30 days), slower revenue ramp, need to hire 2-3 enterprise AEs ($1M cost). What we give up: SMB self-serve velocity. What could go wrong: we lose the SMB beachhead before enterprise lands. Mitigation: keep SMB on lifecycle, dedicate 70/30 split for 2 quarters before fully committing."

**How to catch it.** If your strategy answer has no "trade-off", "what we give up", or "risk" section, it is not a strategy.

---

## Red Flag 8: Metric Selection Without Counter-Metrics

**Symptom.** "I'd measure success by DAU growth." No mention of what could go wrong if DAU grows but other metrics break.

**Why it's bad.** Single-metric optimization is how PMs ship growth hacks that damage retention or NPS. Mature interviewers expect counter-metrics. Skipping them signals product immaturity.

**Bad example:**
> "Success metric: notifications opened per user per week. We want to drive this up."

**Good example:**
> "Primary: notifications opened per user per week. Counter-metrics: notification opt-out rate, app uninstalls, NPS for new users. We optimize the primary only as long as the counters are stable; if opt-out spikes 10%, we stop and re-evaluate."

**How to catch it.** Every metric proposal should include at least one counter-metric. If not, your metrics will incentivize damage.

---

## Red Flag 9: Wrong Level Calibration (APM Answer at Senior Bar)

**Symptom.** Candidate is interviewing for Senior PM but gives PM-level answers — feature-focused, single-team scope, no portfolio thinking.

**Why it's bad.** Each level has a different bar. Senior PM answers need to show scope across teams, multi-quarter horizon, and influence over peers. PM-level answers at Senior bar produce a "good PM, not yet Senior" verdict.

**Bad example:**
> Senior PM interview: "I would A/B test two button colors on the checkout page and pick the winner."

**Good example:**
> Senior PM interview: "I would zoom out — is checkout the right place to invest? In the funnel, where is the biggest activation gap? Then prioritize: 70% of test capacity on the biggest gap, 30% on smaller checkout opts. The button-color test is a tactic; the question is whether checkout is the strategy."

**How to catch it.** Re-read the SKILL.md's level calibration. Does your answer match the *level you are interviewing for*?

---

## Red Flag 10: Pretending to Know Things You Don't

**Symptom.** Interviewer asks about a domain you have no experience in. You bluff.

**Why it's bad.** Interviewers test for intellectual honesty. Bluffing is easily detected and usually fatal. The right move is to acknowledge the gap and reason from first principles.

**Bad example:**
> Interviewer: "How would you think about a B2B procurement workflow?" Candidate (with no B2B experience): "Well, in B2B procurement we know that buyers behave..."

**Good example:**
> "I've worked primarily in consumer, so I don't have direct B2B procurement experience. Let me reason from first principles — in B2B the buyer and user are usually different, the decision cycle is months not minutes, and procurement adds approval gates. My hypotheses would be... and I'd validate them by..."

**How to catch it.** Practice saying "I don't have direct experience, but here's how I'd think about it." Comfort with this phrase predicts interview success.

---

## Red Flag 11: Not Practicing Out Loud

**Symptom.** Candidate has read every PM interview book, written outlines, but has never said an answer aloud or done a recorded mock.

**Why it's bad.** Interview performance is a verbal skill. Reading and writing prepare the structure; only practice aloud reveals timing, pacing, filler words, and how the framework actually plays in real time.

**Bad example:**
> "I've read CIRCLES, I've read Decode and Conquer, I've written 30 outlines. I'm ready."

**Good example:**
> "I've done 12 mocks with peers and 4 with paid PM interviewers. I've recorded 3 and watched myself back. I've identified that I rush the clarifying-questions phase and now consciously slow down. Last 3 mocks I scored above the bar."

**How to catch it.** How many recorded mocks have you done? If under 5, you are under-prepared.

---

## Red Flag 12: Treating the "Any Questions for Me?" Slot as a Throwaway

**Symptom.** End of interview. Candidate asks "what's the team culture like?" — generic, low-signal.

**Why it's bad.** This slot is a test of intellectual curiosity and seniority. Generic questions signal disinterest or lack of preparation. Strong candidates ask questions that reveal they have thought about the role specifically.

**Bad example:**
> "What's the team culture like? What do you like about working here?"

**Good example:**
> "You mentioned the team is moving from feature-team to product-trio model. What's been the hardest part of that shift, and where are you 3 months in? Also — if I joined and 6 months in, the team said I was not living up to expectations, what's the most likely reason?"

**How to catch it.** Write your 3 questions before the interview. None should be Google-able from the company website.

---

## Red Flag Quick Reference

| # | Anti-pattern | One-line check |
|---|---|---|
| 1 | Rote Framework Recitation | Did you say the framework name aloud? |
| 2 | Ignoring Interviewer Signals | Did the next 30s address their interjection? |
| 3 | Missing the "R" in STAR | Numbers + learning in every behavioral? |
| 4 | Opinions as Facts | "Hypothesis" / "validate" / "evidence" present? |
| 5 | Estimation Without Math | Multipliers visible in the breakdown? |
| 6 | Lone-Wolf Hero Stories | "We" outnumbers "I" or roughly balanced? |
| 7 | Strategy with No Trade-Off | "Trade-off" / "what we give up" present? |
| 8 | No Counter-Metric | Every primary metric has a counter? |
| 9 | Wrong Level Calibration | Answer matches level being interviewed for? |
| 10 | Bluffing | Comfortable with "I don't have direct experience"? |
| 11 | No Out-Loud Practice | 5+ recorded mocks? |
| 12 | Throwaway "Any Questions?" | Questions are not Google-able? |

## Related Reading

- SKILL.md Troubleshooting section (for symptom -> root cause -> resolution)
- references/circles-aarm-star.md (for the canonical frameworks, if present)
- pm-career-ladder/references/red-flags.md (for level calibration anchors)
slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack."Stars: 198kUpdated Claude CodeCodexVerifiedInstall
ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report.Stars: 194kUpdated Claude CodeCodexVerifiedInstall
ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt.Stars: 194kUpdated Claude CodeCodexVerifiedInstall