Product Sense Interview Answer
Six-part spine (clarify → rationale → goal → segmentation → pain → solution) for PM product-sense interviews — sound thoughtful out loud instead of over-scripted on paper.
You walked into the "How would you improve YouTube?" interview, brain-dumped four feature ideas, and walked out wondering why you didn't get the offer. Product sense isn't about creativity — it's about reasoning out loud. Establish the user, goal, and pain before any solution, and the answer feels earned. Skip those, and even good ideas sound ungrounded.
Who it's for: PM candidates practicing product-sense interviews, hiring managers running mock interviews, coaches teaching FAANG-style prompts, candidates moving to senior PM roles
Example
"How would you improve YouTube?" → Clarify (creators vs. viewers vs. advertisers?) → Rationale (why now) → Goal (help beginner learners find content they're glad they watched) → Segment (beginner learners vs. enthusiasts) → Pain (no structured progression) → MVP (Learning Paths) with v1 exclusions
New here? 3-minute setup guide → | Already set up? Copy the template below.
# Product Sense Interview Answer
Coach PM candidates through open-ended product-sense interviews using a six-part answer spine: clarify, rationale, goal, segmentation, pain points, solution. Sound thoughtful out loud — not over-scripted on the page.
Not a memorize-and-recite script. A reasoning scaffold that prevents solution-jumping.
## What These Interviews Test
- Clarify ambiguous prompts without getting stuck
- Tie user value to market/business logic
- Segment thoughtfully (not "everyone")
- Prioritize one pain point (not ten equally)
- Make tradeoffs explicit when picking MVP
- Communicate clearly under time pressure
## The Six-Part Answer Spine
1. **Clarify** — reduce ambiguity, define scope, state assumptions
2. **Rationale** — why this matters now (market + company)
3. **Product Goal** — user outcome before features
4. **Segmentation** — who first, why they win
5. **Pain Points** — map journey, name frictions, pick one
6. **Solution** — distinct options, compare, commit to MVP with exclusions
Order matters. Skip from prompt to features → clever but ungrounded. Establish user/goal/pain first → solution feels earned.
## When to Use
**Use:** Product design, product improvement, "build next" prompts, mock interviews.
**Don't use:** Behavioral (STAR), execution/analytics (metrics diagnosis), GTM/pricing.
## Delivery Rules
- State your structure early
- Ask only 1-2 clarifying questions (more = stalling)
- Keep lists MECE (segments distinct, pains non-overlapping, solutions varied)
- Short conversational sentences
- No fake company-mission talk if no company is named — assume startup
## Application
### Step 1: Clarify
Surface the 2 ambiguities that change the answer most:
- Product/surface area
- User group
- Time horizon
- Business model
If interviewer doesn't answer, state assumptions and move on.
**Quality bar:** Materially-changing questions. "Optimize for viewers, creators, or advertisers?" beats "mobile or desktop?"
### Step 2: Rationale
**Market view:** big/strategic? real pain? why now?
**Company (if named):** mission fit, business objective, competitive landscape, gap, unique strength.
End with one-line thesis. Use qualitative signals unless you know numbers cold.
### Step 3: Product Goal
`Help [user] [achieve outcome], so that [broader impact].`
✅ "Help beginner YouTube learners find content they're glad they watched, so the platform becomes an intentional learning destination."
❌ "Build a personalized AI learning path feature." (solution disguised as goal)
### Step 4: Segmentation
Identify ecosystem players → choose target → pick 2 dimensions that change needs:
- Goal/JTBD
- Stakes/consequence level
- Expertise level
- Workflow constraints
- Frequency
Avoid demographic cuts that don't change the product.
After choosing target: brief reach/impact/strategic-fit rationale + 2-sentence persona (no pain points yet).
### Step 5: Pain Points
Break user journey into 4-6 stages. List frictions. Prioritize the top one using:
- **Frequency** — how often
- **Severity** — how blocking, how underserved, emotional cost
This is the fulcrum. Vague pain → generic solution.
**Quality bar:** Pain ≠ missing feature. "No structured progression after each video" is a pain. "No AI learning path" is already a solution.
### Step 6: Generate, Choose, Close
Three distinct solutions (different mechanisms, not feature line-items). Evaluate on user impact + effort. Commit to one MVP with:
- Core features
- 1-2 explicit v1 exclusions
- Top risks + mitigations
Close with one sentence: target segment + top pain + first bet. That's what the interviewer remembers.
## Anti-Pattern Example
"I would improve YouTube by adding AI summaries, better recommendations, creator analytics, and a study mode."
Why it fails:
- No target user
- No prioritized pain
- No business logic
- Four ideas never compared
## Common Pitfalls
- **Solution-first** — write goal + pain before brainstorming
- **Segmentation theater** — explicit reach/impact/fit comparison
- **Goal-as-feature** — rewrite as user outcome
- **Pain points that are solutions** — user-language first
- **Three fake options** — vary the mechanism, not packaging
- **Weak close** — restate target + pain + first bet
- **Over-answering** — focus over breadth
## References
- `problem-statement`
- `proto-persona`
- `customer-journey-map`
- `opportunity-solution-tree`
- Lewis Lin, *Decode and Conquer*
- Gayle Laakmann McDowell & Jackie Bavaro, *Cracking the PM Interview*
Get new playbooks like this one
One email a week with new Claude Code workflows. Free, like everything here.
No spam. Unsubscribe anytime.
What This Does
Coaches you through 6 sequenced sections (clarify → rationale → goal → segmentation → pain → solution) with quality bars at each step. Prevents solution-jumping, segmentation theater, and weak closes. Outputs an answer that sounds conversational, prioritizes one user/pain/MVP, and ends with a memorable one-sentence recap.
Pairs with problem-statement, proto-persona, and opportunity-solution-tree for deeper practice.
Quick Start
mkdir -p ~/Documents/ProductSenseInterview
mv ~/Downloads/CLAUDE.md ~/Documents/ProductSenseInterview/
cd ~/Documents/ProductSenseInterview
claude
Provide the prompt ("How would you improve X?", "Design Y for Z", "What would you build next?"). Claude coaches you through the six-part spine and stress-tests the answer.
The Six-Part Spine
| # | Section | Output |
|---|---|---|
| 1 | Clarify | 1-2 scope-changing questions or stated assumptions |
| 2 | Rationale | One-line thesis (market + company logic) |
| 3 | Product Goal | Help [user] [outcome], so that [impact] |
| 4 | Segmentation | Target segment + 2-sentence persona |
| 5 | Pain Points | Journey map + one prioritized pain |
| 6 | Solution | 3 distinct options + chosen MVP + exclusions + closing recap |
Tips & Best Practices
- State structure early. "I'll clarify, then walk through goal, segment, pain, and MVP" — interviewer can follow.
- Cap clarifiers at 2. More than that feels like stalling.
- One pain, not three. Severity + frequency lens. The fulcrum of the whole answer.
- Three distinct solutions. Different mechanisms, not three flavors of the same idea.
- Memorable close. "For beginner learners, the top pain is no structured progression, and my first bet is Learning Paths." That's what gets repeated to the hiring panel.
Common Pitfalls
- Feature dump before naming user/goal/pain
- Listing 5 segments and choosing one with no rationale
- Goal that's actually a feature ("build an AI learning path")
- Pain points that are missing features ("no AI summary")
- Three "solutions" that are one solution with different packaging
- Ending after listing features (no closing recap)