Product sense interviews are where PM candidates either shine or expose that they're not actually PMs yet.
The questions seem simple: "How would you improve Instagram?" or "Design a product for X." But they're testing something that's hard to fake—whether you actually think like a product person.
I've seen engineers with perfect resumes bomb these. I've seen career changers with limited experience nail them. The difference is usually not intelligence or preparation. It's whether you genuinely orient around users and problems, or whether you think in features first.
What Product Sense Actually Means
Let's define the thing before we figure out how to demonstrate it.
Product sense is the ability to:
- •Identify real problems worth solving (not just interesting ones)
- •Understand user needs at a level deeper than what users say they want
- •Generate solutions that elegantly address those needs
- •Evaluate tradeoffs between different approaches
- •Prioritize ruthlessly based on impact and effort
- •Anticipate consequences of product decisions
The product sense interview is designed to test all of these in 30-45 minutes. No pressure.
The Types of Product Sense Questions
Most product sense questions fall into a few categories:
Product Improvement "How would you improve [existing product]?" "What would you change about [feature]?"
New Product Design "Design a product for [user segment] to [solve problem]" "Build a [type of product] for [company]"
Favorite Product "What's a product you love? Why?" "Tell me about a product that's well-designed."
Product Critique "What's a product that's poorly designed? How would you fix it?" "What would you change about our product?"
Each type is testing slightly different things, but the underlying question is the same: do you think like a PM?
The Structure That Works
I'm going to walk through a framework, but I want to be clear: the framework is scaffolding. It keeps you organized. But the quality of your answer comes from your actual thinking, not from hitting every step.
Step 1: Clarify (30 seconds)
Before you dive in, make sure you understand the question and align on scope.
"Before I start, a few clarifying questions. When you say 'improve Instagram,' should I focus on a specific user segment or the whole product? Should I assume we have Instagram's full resources, or are there constraints I should consider?"
This does two things: it shows you don't make assumptions, and it gives you useful information to work with.
Don't ask too many clarifying questions—two or three is enough. You're showing judgment, not stalling.
Step 2: Frame the Problem (1-2 minutes)
Start with what you understand about the product, users, and company.
"Let me start with how I think about Instagram. The core user segments are content creators and content consumers—with a lot of overlap. The product serves different jobs: some people come to kill time, some to stay connected with friends, some to discover new content, and some to build an audience.
Instagram's business model is advertising, so engagement metrics matter. But they're also competing with TikTok for attention, especially among younger users.
Given that context, I want to focus on [specific area] because [reasoning]."
This framing shows you understand the product landscape before proposing solutions. It's the foundation everything else builds on.
Step 3: Identify Problems (2-3 minutes)
Now dig into specific problems worth solving. Don't just list features you'd add—start with user pain points.
"For content creators specifically, I see a few key problems:
First, discoverability. New creators struggle to get their content seen. The algorithm favors established accounts, which makes it hard to build an audience from scratch.
Second, feedback loops. Creators don't get enough actionable information about what's working. Analytics exist but they're not very useful for improving content.
Third, monetization. Compared to YouTube or TikTok, Instagram's monetization options for creators are limited.
I'd prioritize the discoverability problem because it's upstream of the others—if you can't get views, nothing else matters."
Notice the structure: multiple problems identified, clear prioritization, and a reason for the prioritization.
Step 4: Generate Solutions (3-4 minutes)
Now propose solutions to your prioritized problem. Offer 2-3 options with tradeoffs.
"For the discoverability problem, I'd consider a few approaches:
Option A: A new creator boost program. Guarantee first-time creators a minimum number of impressions on their first few posts. This removes the 'posting into the void' anxiety and gives the algorithm enough signal to judge content quality.
Option B: Curated 'rising creators' sections. Add a dedicated feed or tab surfacing creators under a certain follower threshold who are posting quality content. Let viewers opt into discovering new voices.
Option C: Collaboration matching. Help new creators collaborate with established ones through some kind of matchmaking feature. This gives new creators access to existing audiences.
Each has tradeoffs. Option A is highest impact but might flood the feed with low-quality content. Option B requires curation investment. Option C is the most complex to build.
I'd lean toward Option A because it addresses the core problem most directly with the simplest implementation. We could mitigate the quality concern by tying the boost to early engagement signals."
Step 5: Define Success (1 minute)
How would you know if your solution worked?
"To measure success, I'd look at:
Primary metric: Percentage of new creators who post three or more times in their first month. This is an activation metric—are we getting people past the initial hurdle?
Secondary metric: 90-day creator retention rate.
Guardrail metric: Overall feed engagement rate—we want to make sure we're not degrading the consumer experience by surfacing lower-quality content.
I'd run this as an A/B test with a cohort of new accounts over 6-8 weeks."
Step 6: Acknowledge Limitations (30 seconds)
Show self-awareness by acknowledging what you don't know.
"A few things I'd want to validate: Is the discoverability problem actually the biggest pain point for creators? I'm hypothesizing based on what I know, but I'd want to talk to creators directly. And I'd want to understand the technical constraints around algorithm changes—there might be reasons this is harder than it sounds."
This shows intellectual honesty, which interviewers appreciate.
The Most Common Mistakes
Jumping to solutions too fast. The first three minutes of your answer should be about understanding, not proposing. If you start with "I'd add a feature that..." you've skipped the most important part.
Picking the wrong problem. Some problems are more important than others. If your proposed improvement doesn't address something that actually matters to users or the business, you've missed the point.
Feature vomiting. Listing ten features is not impressive. Deeply thinking through one or two is. Depth beats breadth.
Ignoring business context. A consumer social app and a B2B SaaS product have different goals. Your answer should reflect that you understand what success means for this particular product.
Being too generic. "I'd improve the user experience" means nothing. "I'd reduce the steps from discovery to first purchase by 40%" means something. Be specific.
Not having opinions. "Well, it depends..." is a weak start. Take a position. You can acknowledge uncertainty, but have a point of view.
How to Prepare
Build product intuition (ongoing)
The best preparation for product sense interviews is being genuinely curious about products all the time. Use apps critically. Notice what works and doesn't. Think about why products make the decisions they do.
If you're not doing this naturally, start. Pick an app a day and spend 10 minutes thinking about what you'd change and why.
Study the company's product (before each interview)
If you're interviewing at Spotify, use Spotify intensively for a week before. Note pain points. Observe what seems intentional versus what seems broken. Have opinions ready.
You might get a question about their specific product. Or you might not. Either way, having depth on their product helps.
Practice out loud (before interviews)
Product sense is a verbal exercise. Get comfortable talking through your thinking. Practice with friends, record yourself, do mock interviews.
The thinking and the communication are both being evaluated. Practice both.
Know common products deeply
Instagram, Netflix, Uber, Spotify, Amazon—these come up a lot. Have ready answers for "how would you improve X" for the big consumer products.
But have ready answers, not scripted ones. Know them well enough to adapt based on clarifying information.
Sample Questions to Practice
Easy to medium:
- •How would you improve the Netflix home screen?
- •Design a product to help people learn a new language.
- •What's a product you think is poorly designed? How would you fix it?
- •How would you improve Uber for drivers?
Harder:
- •Design a product for elderly people to stay connected with family.
- •How would you improve LinkedIn for job seekers?
- •Design a fitness product for people who hate working out.
- •What product would you build if you joined our company?
The hardest product sense questions are about the company's own product—because you're expected to have done your homework, and they know the constraints better than you do.
The Meta-Point
Here's what really separates good product sense answers from great ones:
Great answers feel like a conversation with someone who genuinely cares about solving the problem. Not a presentation. Not a framework walkthrough. A thoughtful person working through a hard problem out loud.
The framework keeps you organized. Your actual curiosity about users and products is what makes the answer good.
If you don't have that curiosity naturally, product sense interviews will always feel like a performance. If you do have it, the interview is just an opportunity to show how you think.
Related reading: Behavioral Interview Questions and Analytical Interview Questions for the other major PM interview types.