What they're testing
Product sense interviews are 45 minutes. You get ~35 minutes of actual content time. Top companies like Meta, Stripe, Airbnb, Doordash, Lyft use them because they're the closest simulation of the real job: turn ambiguity into a prioritized, user-validated bet.
Interviewers score across five dimensions, and you need to be solid in all of them:
1. User empathy: Can you get inside a user's head and not project your own preferences?
2. Structured thinking: Can you bring clarity to a vague problem?
3. Prioritization: Can you make a defensible trade-off call?
4. Business acumen: Do your solutions make strategic sense for the company?
5. Creativity: Do your ideas go beyond the obvious?
Weak in any one of these and you don't make it through, regardless of how strong you are in the others.
The three question types
Product sense is an umbrella term. Know which type you're dealing with so you pick the right lens.
1. Product design (~40% of questions): Design something new from scratch.
"Design a product for elderly users in VR." "Design a pen for an astronaut."
2. Product improvement (~40% of questions): Improve an existing product.
"How would you improve Facebook Groups?" "How can Airbnb increase bookings?"
3. Product strategy (~20% of questions): Broader company-level thinking.
"What should Meta do next?" "What product would you build to expand into fitness?"
The framework that works
Forget memorizing a rigid acronym. The goal is structured thinking, not reciting steps. Here's the flow that holds up across all three question types:
Step 1: Clarify before you commit (2-3 min)
Take 20 seconds before you say a word. Write down the prompt. Then ask targeted questions to narrow scope. Not a generic "tell me more" opener. Ask specific questions like:
• "Are we targeting existing users or net new?"
• "Is this mobile-first or cross-platform?"
• "Are we optimizing for growth, retention, or monetization?"
State your assumptions out loud to the interviewer, who will course-correct any drift. Constraints are useful as they focus your design.
Step 2: Define the mission and goal (1-2 min)
Before users, before features you need to anchor to purpose. What is the company's mission? Why does this product area matter to their strategy? This tells the interviewer you think like a PM, not a designer.
"Airbnb's mission is to create a world where anyone can belong anywhere. An improvement to their search experience should ladder up to trust and discovery, not conversion rate alone."
Step 3: Segment your users (3-4 min)
List broad user segments, then choose one as your focus. Don't design for everyone as that's how you build something nobody loves.
Use two filters when prioritizing your segment:
• Strategy alignment: Does targeting this segment fit the company's strategic goals?
• Strength alignment: Does the company have the right capabilities for this segment?
Once you've picked a segment, build a quick persona. Make it real. Give them a name, a job, a specific frustration. "My friend Jake is a 34-year-old software engineer with a newborn. He's awake at 2am, one hand on a bottle, trying to book a last-minute weekend trip." That specificity makes you memorable and shows genuine empathy.
Step 4: Map the user journey (3-4 min)
Walk through the experience step by step before identifying pain points. This is where most candidates skip ahead and pay the price. A user journey forces you to find actual pain rather than assumed pain.
Example for a travel app: Search → Compare → Select → Book → Pre-trip → In-trip → Post-trip
Flag the friction points along the way. The best problem statements come from this exercise.
Step 5: Identify and prioritize pain points (2-3 min)
List 3-5 pain points from the journey. Then explicitly prioritize a single pain point. State your prioritization criteria out loud, like frequency, severity, or uniqueness, instead of leaving them an unstated assumption.
"I'm going to focus on the booking abandonment moment — it's high-frequency, high-intent, and the current experience introduces unnecessary friction right when the user is closest to conversion."
Step 6: Brainstorm solutions (4-5 min)
Generate 3-4 solutions. Don't self-censor during brainstorm. Include at least one "moonshot" idea, a transformative idea, not just an incremental feature tweak. This shows range. Then narrow the list to 1-2 solutions for further development.
Apply the same strategy and strength alignment check you used in user segmentation. A clever idea that doesn't fit the company's roadmap or capabilities isn't a good answer.
Step 7: Define success metrics (2-3 min)
Pick metrics that are specific and tied to your targeted pain point. Avoid vanity metrics.
• North Star: What's the primary outcome? (e.g., bookings completed per session)
• Leading indicators: What early signals show you're on the right track? (e.g., time-to-book, abandonment rate at payment step)
• Guardrails: What's the acceptable downside? (e.g., cancellation rate, support tickets)
Step 8: Summarize and tie it together (1-2 min)
Close with a clean summary. "We designed X for Y users to solve Z pain point, so that [outcome for user] and [outcome for business]." One sentence. Make the interviewer nod.
The mistakes that kill candidacies
Jumping straight to features. The most common and most damaging mistake. If you're pitching features within the first 3 minutes, you've already lost the room. The interviewer is checking whether you understand the problem before you jump to solutions.
Designing for every user. Vague users produce vague solutions. A sharper target segment buys your design more credibility and differentiation.
Using frameworks as a crutch. Reciting CIRCLES steps verbatim signals you studied for the test, not the job. Frameworks are scaffolding, not a script. Make the conversation feel like a natural exchange. Check in with the interviewer. Adapt based on their cues.
Inconsistent prioritization logic. You pick user segment A based on criteria X, then you pick a feature that serves user segment B. Interviewers catch that inconsistency. Use the same prioritization logic from start to finish.
Forgetting the business. This is a PM interview, not a UX interview. Every solution needs to connect back to why it makes strategic sense for the company.
How to practice
You need reps. Here's a focused prep plan:
1. Build a product teardown habit. Every app you use, ask: who's the target user? What pain does this solve? What would you change, and what's the reasoning? This sharpens intuition faster than studying frameworks.
2. Practice out loud. Your answers sound very different in your head vs. out loud. Record your answers. Listen back to the recording. You'll catch the rambling, the rushed moments, and the gaps in your logic.
3. Run timed mocks. 35 minutes is a tight window. Knowing the pace of each step, when to move on, when to go deeper, only comes from practice under time pressure.
4. Use real company prompts. "Design a product for Meta to enter healthcare" is harder and more useful than generic prompts. Find current-year question banks (Lewis Lin's is consistently updated) and work through the questions.
5. Master 2-3 companies in depth. Know their mission, their current product bets, their competitive pressures. If you're interviewing at Meta, you should be able to connect any product design question back to their AI strategy and their ad revenue model.
The mindset shift that makes the difference
The candidates who land the offer treat the interview as a product conversation, not a test. They're genuinely curious about the problem. They think out loud, naturally and without a script. They say, "that's an interesting constraint, it changes my approach," and they mean every word.
The interviewer has done this exercise dozens of times. They can tell the difference between someone who memorized a framework and someone who thinks in products. The goal is to be the second person.
Product sense can't be faked long-term, but it absolutely can be developed with practice. Build the habit of thinking this way about every product you touch, and the interview becomes a natural extension of how you already see the world.